AID EVALUATION PAPER 3

AID EVALUATION PAPER 3

EXTENSION IN THE ANDES

An Evaluation of Official U.S. Assistanceto Agricultural Extension Services

in Central and South America

Condensed

by

E.B. RicePPC /Evaluation Staff

April 1971

AID Evaluation Papers represent the views of their authors and arenot intended as statements of Agency policy.

2

In 1967, the national extension service ofColombia was shifted to the control of thenational research service. A bean-breederwas put in charge. Extensionists cite thatdate as the end of effective extension inColombia. Other professional agriculturalistscite it as the beginning.

3

FOREWORD

Two nearly universal problems for development administra­tors in less developed countries are to adapt institlJ.tions so thatthey can cope effectively with local problems and to increase lowagricultural productivity.

This evaluation report deals with both problems in a com­parative study covering a 30 -year experience with extension ser­vices in twelve Latin American countries. Some comparisons arealso made with four other countries. Not only is the report thorough,but it is readable. It contains a flavor of the officers involved, bothLatin and Yankee, and presents numerous village and case studies.

The findings about institution building, about the methods andimpact of extension services, and about a strategy for raisingagricultural productivity are general to all countries studied.Therefore, the conclusions of the author should be useful to practi­tioners in development throughout the world, whether these prac­titioners are employed by developing countries or by donor assistanceagencies, public and private.

The author, Edward B. Rice, has been a member of theEvaluation Staff of the Office of Program and Policy Coordination ofthe Agency for International Development for the past three yearsand has had overseas experience in both public and private employ­ment. As a member of the Evaluation Staff, Mr. Rice played amajor part in the preparation for the AID Administrator's SpringReviews on The New Cereal Varieties and Land Reform.

This study has been reviewed by Professors Richard S. Eckaus,Everett E. Hagen and Matthew Edel of the Department of Economicsof the Massachusetts Institute of Technology, and by the distinguishedpersons whose critiques are appended. It has also been read byvarious officers of the Agency for International Development and ofthe U. S. Department of Agriculture. All approve of the rigorousnature of Mr. Rice's data collection and analysis. His conclusionsdeserve attention. As with other papers in this series of evalua-tion reports, however, the conclusions are those of the author and notof the Agency.

C. William KontosDirector of Program EvaluationAgency for Internationai Development

5

PREFACE

There are two audiences for the general report, a dissertationcommittee and a community of program officers. I have assumedthat it is possible to design an evaluation study that is neither toooperational for the one, nor too academic for the other. At leastthat was my intention. I have hoped also to have the study serveseveral purposes for the Agency for International Development:first, as an analysis of an important part of its rural developmentstrategy; second, as a useful design for learning lessons abouttechnical assistance through retrospective evaluation; third, as anexample of the fruitful relations that AID could develop with otherdissertation committees.

This paper is an abridged edition of that report. The conden­sation process treated the sections of the report unevenly. Theprincipal economic analysis, of the impact of extension on farm pro­ductivity (a section which occupies half of the general report), isrecapitulated here in one chapter (4). The description of the U. S.programs, and the analysis of institution building, have been cutmuch less (chapters 2 and 3). The other three sections -- theintroduction, the discussion of efficiency, and the conclusions --are repeated almost in their entirety 0, 5 and 6). What have largelydisappeared are the factual material and analyses which supportmany of the assertations made in this version. The author is notso foolish as to pretend that he "prove$' his points: the field of in­quiry is too uncertain for that. But he invites the reader to referto the general report, to see whether the mass of evidence presentedthere allows any other set of conclusions.

My debts are many. The earliest are to the Department ofEconomics of the Massachusetts Institute of Technology, for itspatience, and to Professor Richard S. Eckaus, for helping strengthenthe analysis of productivity. Next are my Lo.u. 's to the Bureau forProgram and Policy Coordination in AID, and to three consecutivechiefs of the Evaluation Staff who permitted me to mix self-interestwith company business. There is the obligation to Joseph L. Matthews,Assistant Administrator for International Extension in the (Federal)Extension Service, who encouraged me in the beginning, helped mein the middle, and was still a friend in the end.

Milo L. Cox and Donald L. Fiester of the Bureau for LatinAmerica's agricultural office gave me continuing consultation onsubject matter, and introduced me to theUSAID agricultural staffswithout whose help the study could not have been completed. A few

6

on those staffs, including Alphonse C. Chable, Earl Threadgould,Richard L. Hughes and Ralph L. Van Dixhorn, were hit harderthan others, but I am indebted to all who shared with me theirmemories, thoughts and time.

As I review my interview notes, I realize the extent to whichmy own conclusions reflect the concerns of Milton J. Lobell" pres­ently the USAID Food and Agriculture Officer in Paraguay. I thinkmy several talks with him mattered more than most.

Among the U. S. extension advisers now repatriated toWashington, Nazario C' de Baca and Ernest N. Gutierrez providedinvaluable glimpses of the way it was in the golden years, and I hopemy efforts in the text to disassociate my remarks about the overallprogram from my feelings about their individual efforts have notbeen missed.

This sort of disassociated appreciation is owed to the manyhost country extension officers who helped me everywhere I went.I should list all the extension directors, since each went out of hisway to provide the information I requested. The one who helpedthe most, Carlos A. Anleu, is likely to be most annoyed, but hehas learned to be philosophical about AID "findings" and I supposealready suspects that my own findings will be upended by the nextU. S. evaluation officer who looks at his program. To the regularextension headquarter staffs, I must add the FERTICA organiza­tions in Guatemala and EI Salvador, the Caja Agraria personnel inBoyaca, Oscar Roca in Cochabamba, and Juan Calderon in La Paz.The success of my field work depended upon them. I would like tosingle out for special mention Humberto Rosado, who was my firstand most important contact with the Latin American extension com­munity, and offered his files at the IlCA office in Guatemala Cityfor my research.

A number of persons reviewed the draft and helped me improveit. Among those persons, Robert L. Hubbell and Joseph E. Walkerof AID, and Remy T. Freire of OAS, went beyond the call of duty,giving me the kind of detailed comments which often save an authorfrom a series of embarrassing stumbles.

Several busy secretaries in AID are glad this is over, especi­ally Marion Boch, Joyce E. Williams, Lucille W. Stanton andShirley A. Gigliell.

7

I have a special debt to Joseph Di Franco. now with FAD inRome, whose philosophy, presence, and impact pervade the subjectmatter of this study. I refer to him frequently, although I havenever met him. Some of his earlier associates tell me I have beenunfair, on occasion, that I have taken advantage of a man for exposi­tional purposes.

Finally, my thanks to my mother, Persia Campbell, whosehelp and inspiration were decisive. and to my wife, Liz, withoutwhom this wouldn It have been such fun.

EXTENSION IN THE ANDE S

(Condensed)

35

11

Page

Foreword, by C. William KontosPrefaceNotes on Terminology and Abbreviation

TABLE OF CONTENTS

2. U. S. PROGRAMS IN EXTENSIONA. 1940-1950 29B. 1950-1958 31C. 1958-1970 37D. Financial Assistance 41

3. EFFECTIVENESS: BUILDING EXTENSIONINSTITUTIONSA. Appropriateness of the Study 43B. Extension Organizations 44C. Application of the Criteria for Success 46D. Adoption versus Adaption 61E. Education versus Fomento 62F. Comparisons between Countries within the

Study Area 65G. Comparison with other Countries 66

4. SIGNIFICANCE: INCREASING AGRICULTURALPRODUCTIVITYA. Productivity and other Significance Effects 71B. Design of the Productivity Study 73C. Findings 83

1. INTRODUCTIONA. Agricultural Extension 14B. Study Design 17C. Research Program 20D. Study Area 26

Chapter

CRITIQUESA. J. L. Matthews, USDA/Extension ServiceB. Remy Freire, OAS/Department of Economic AffairsC. A. T. Mosher, Agricultural Development CouncilD. Delbert T. Myren, AID/Office of AID Research and

University RelationsE. Remarks by the author

APPENDIXA. Maps of the Study Districts 123B. Forms

1. Director's Worksheet 1292. Professional Questionnaire 135

C. Related Issues in Institution Building1. Regional Demonstration Programs and

Local Governments 1412. Servicios as an Instrument for

Institution Building 1443. U. S. Extension Advisors 149

Page

109117120

97101104

6. CONCLUSIONS AND PRESCRIPTIONSA. ConclusionsB. PrescriptionsC. Final Remarks

5. EFFICIENCY: COSTS, BENEFITS ANDALTERNATIVESA. CostsB. BenefitsC. Alternatives

Chapter

11

Notes on Terminology and Abbreviation

1. Unless otherwise modified, the word American refers tocitizens of the U. S.

2. The Inter-American Development Bank, commonly referredto in AID as IDB or IADB, is referred to in this paper by itsSpanish initials BID (Banco Interamericano de Desarrollo).

3. The Inter-American Institute for Agricultural Sciences, com­monly referred to in AID as lIAS, is referred to in this paperby its Spanish initials I1CA (Instituto Interamericano deCiencias Agricolas). I1CA's principal training institute is atTurrialba, Costa Rica, and when the Latin American agricul­tural community refers to Turrialba, it usually refers to theI1CA training institute.

4. The text indicates whether the initials ICA refer to the Inter­national Cooperation Administration (AID's predecessor) orto the Instituto Colombiano Agropecuario (Colombia's research­extension authority).

5. The Federal Extension Service of the U.S. Department ofAgriculture (USDA) changed its name in 1970 to the ExtensionService. The old name and its initials, FES, are used in thispaper. The FES is the federal agency which backstops theindependent state extension services, and, with them, formsthe Cooperative Extension Service.

6. When AID contracts another U. S. Government agency, such asthe USDA, to staff AID financed technical assistance programsthe arrangement is called a Participating Agency Service Agree­ment, or PASA.

7. Extension service refers to the national system, whereas exten­sion agency refers to an individual field office. The word agencyis used here only when referring to a field extension office,except when it refers to the Agency for International Development.

8. Servicio has a special meaning in AID, where it refers to a jointoperational authority. But is a Spanish word which simply meansservice, and Latin Americans usually don't know what AID isdriving at when it asks them to comment on the servicio problem.

12

9. Fomento organizations are developmental authorities whichcarryon action programs as well as give information andadvice.

10. Some Latin American countries use the manzana (1.70 acres)as a unit of land holding. Others use the hectare (2.47 acres).Both are used in this paper. along with acre. depending uponthe units used in the prime information source. The quintal(100 lbs) is used in many countries as a measure of weight.

11. IB refers to a particular model of the institution-buildingprocess. (See chapter 3. section c).

12. The accent mark is omitted from Spanish words throughoutthe report.

13. The words mission, and USAID, refer to AID's overseasoperational staffs.

CHAPTER 1

INTRODUCTION

This is a study of the results of thirty years of U. S. effortsto help build agricultural extension institutions in Latin America.The subject has intrinsic interest, and also offers a base for ex­ploring some critical issues associated with technical assistance,for a number of reasons. First, institution building is consideredby most U. S. foreign aid experts to be the most enduring phase oftechnical assistance activities, 1 and we have a chance here to seeif it worked on an important occasion. Second, agricultural exten­sion as conventionally described is an American product, and exten­sion services did not exist in most of Latin America before theAmerican technical teams arrived. We have, therefore, a goodexample of cross -cultural "transfer". Third, American technicalassistance programs are older in Latin America than in other partsof the world, agricultural programs are the oldest of all, and exten­sion has always been a part of the American program, and, duringthe 1950s, was the centerpiece of most U. S. rural developmentcountry strategies. If we had to select one technical assistance

lIn his opening remarks to the Conference on Institution Build­ing and Technical Assistance, sponsored by AID in Washington inDecember, 1969, Joel Bernstein, Administrator of the TechnicalAssistance Bureau of AID and co-chairman of the Conference, stressedthat very point:

"Building improved institutional capability shouldtherefore be the main purpose of technical assistance.For the most part. the value gained from varioustechnical assistance efforts is not what the individualprofessional does in the developing country; but rather.it is the expanded knowledge base· and the improvedfunctional capabilities which he leaves behind that aresignificant. "

D. Woods Thomas and Judith G. FeI)der (eds. ) Proceedings:Conference on Institution Building and Te~hnicalAssistance (Washington,D. C., Dec. 4 and 5, 1969; Washington: AID and the Committee onInstitutional Cooperation, 1970), p. 6.

13

14

program to represent the overall U. S. experience, this programwould be a candidate. Fourth, and finally, ten years have passedsince the Americans started phasing out of extension programs inLatin America, a period long enough to allow us now to look forpermanent results. There are few instances of such a long gesta­tion period in the history of major institution building activities.For these reasons the paper is offered as a case study of technicalassistance, although the discussion is limited to extension and noattempt is made before the final paragraphs to read from the resultsthe wider implications.

A. Agricultural Extension

Agricultural extension refers to an organized process ofextending agricultural information; from whatever its source, tofarmers who can use it, through channels other than the formalschooling system. Extension is conceived as supplementing thelatter: reaching farmers who no longer are or never were voca­tional students. In developing countries, where few farmersparticipate in vo -ag studies, extension is by definition the onlychannel available to the majority for getting advice that they can Itget from their neighbors.

One must distinquish between the extension function and theextension institutuon. The function can be assumed by any organ­ization that has an interest in having farmers improve their prac­tices, incomes, or welfare. Fertilizer dealers, tobacco companies,rural credit agencies, etc., are each engaged in getting specializedinformation to clients, suppliers and debtors. Extension institutions,those that are referred to in this study as regular or conventionalextension services, have a special role. They are not selling, buy­ing, or collecting anything: at least they are not supposed to.Conceptually, their job is limited to information delivery. In theU. S. during the early years of the 20th century, and in all developingcountries presently, extension services have also tried to motivatefarmers to take advantage of available information and to inspirethem by establishing successful demonstrations. To that extent,extension institutions playa manipulative role in development. Butthe intrusion is described as benevolent and there is not supposed tobe any interference with the farmer's freedom of choice, or anysubstitution of predetermined goals for self-determined goals.- Atleast, that is the approach professed by the flextension school".

15

The role of the agricultural extension service in a farmcommunity is, in fact, just that: a service to the farmer. Thereis no role for such a service, of course, if farmers refuse it.There is no role for such a service if there is no useful information.There is no role for such a service if the extension function is beinghandled more efficiently by other organizations without seriouslyprejudicing the farmer's self-interest. In the 1940s and early 1950s,when the U. S. overseas extension programs were gathering momentum,it was clear to the architects of these programs that (1) farmers indeveloping countries would not resist if approached in the right way,(2) there was a large amount of useful information that was not beingput to practice, (3) no other organizations were performing the func­tion, and (4) most of the farmers were without political power or arepresentation in government that could look out for their interests.

The assumption that useful information was available to extenddid not go unchallenged. An opposing school of thought, associatedwith the U. S. Department of Agriculture in the days when the USDAand the Institute for Inter-American Affairs were quarreling overthe proper shape for U. S. rural development programs for the 1950s,held that the supposedly available technology was of U.S. origin andunsuited for other lands and peoples: that technologies appropriatefor Latin America had still to be developed. For expositional pur­poses, the issue is described here as a clash of diametrically opposedviews. The differences were probably only over the degree of suita­bility of the information. The opinion that prevailed was that diffusionrather than research was the principal bottleneck, and that therewould be a direct and potentially large payoff to extension itself. Theextension impact would of course be increased by additional research,credit and other inputs, but none of these were essential to extensionfor quick and impressive results.

So the decision was made to build and institutionalize extensionservices. We are asking in this study if the decision, and the programit generated, were correct. Was there a role in a production orientedrural development program for an extension service, with an organiza­tionallife independent of other rural development agencies? Wouldthe agents of the extension service offer the farmers useful informationthey did not already have?

Extension services are more elaborate than the pure diffusionagency described. In the U. S., relationships with research andformal educational services are emphasized and well defined. Ex­tension is given equal footing with these other two professions in the

16

"three legged" land grant university system. Relationships withbanks, soil conservation agents, private dealers and other ruraldevelopment agencies have been tightened. Financial supportderives by formula from all levels of government, and from privateorganizations. The methodology of communication has been improved,and extension education courses at professional schools where futureextension agents are trained expose the students to a variety of infor­mation delivery systems, and teach them the circumstances underwhich each works best. Field demonstrations of superior farm prac­tices and of their results have grown to be the cornerstone of exten­sion activity. Moreover, the philosophy of extension has expandedfrom the original agricultural base, and the role of extension is seento be as much one of strengthening families and rural societies as itis in improving farm practices.

Indeed there is reaSon to argue that the "conventional model"of an extension service presented earlier is not and has never beenrepresentative of extension in the U. S. The job of the county agentshas always been la~er than information delivery. Most agents havedone what they needed to do, including helping farmers build roadsand procure supplies. County agents also serve to inform the re­search staff about the experiments the farmers would like to haverun - - extension in reverse if you will. This alleged interaction pro­duces a dynamic process quite different than an organization to teachfarmers from a static store of information. Purity was never a partof the system replicated in the counties throughout the U. S. Accord­ing to J. K. McDermott, of the Office of AID Research and UniversityRelations, "the purity aspect has been a latter day movement corre­lated with the effort to build professionalism into extension. It movedslowly in the U. S. However, professionalism caught or very rapidlyin the more favorable environment of Latin America". The conven­tional model refers then to the conceptual structure which the Americanprofessional extensionists offered to those countries which appearedready to adopt the institution. It differs from the historical U. S. model,and does not adequately describe the typical U.S. state service, thoughemployees of that service may admire it.

Since this study refers frequently to the "conventional model",and to the "conventional system", the reader is asked to bear in mind

2Memorandum to PPC /PDA, Ted Rice (Washington: AID, March 5,1971), pp. 1-2.

17

that the word conventional has a special meaning: that it refers toa popular conceptual construct, and not to historical experience.For more background on the extension system, the reader shouldlook to other texts. One of them, by Arthur T. Mosher and en­titled Varieties of Extension Education and Community Develop­ment3 is unexcelled in laying out the arguments for and againstthe model. 4

It should be understood that AID has largely abandoned theconventional approach and that we are evaluating here an assistanceprogram that is over.

B. Study Design

The study follows the typology of evaluation studies developedfor AID in 1962 by David Mayer. He distinguished four categoriesof evaluation studies. One is the evaluation of project compliance,which asks whether .legal and administrative requirements weremet. The second i.s'the evaluation of project effectiveness, whichasks whether the project was completed, in the sense of accom­plishing its immedi~te, and usually explicit, targets. The thirdis the evaluation of project significance, which asks whether theproject achieved the higher order objectives of the developmentprogram. The fourth is the evaluation of project efficiency, whichasks whether the immediate targets, or the higher order objectives,could have been reached at lower costs by other means. Thisterminology is now widely used within the Agency. 5 It differs insome respects from those used at the UNDP and other organizations.

3Comparative Extension Publication Number 2; Ithaca:Cornell University, Rural Education Department, 1958.

4To appreciate the sort of extension work which was com­mon practice in the U. S., and which differed in important respectsfrom the idealized model developed in the 1950s to represent it,one can refer again to Mosher, who describes it, and also to othersfudies, for example Fifty Years of Cooperative Extension inWisconsin, 1912-1962, by the University of Wisconsin's ExtensionService (Circular 602; Madison: University of Wisconsin, 1962).

5See the Agency's new Evaluation Handbook (Washington:AID, 1970), p. 5.

18

The difference between effectiveness and significance isfundamental to the typology. For most projects it is possible tofind a narrow set of criteria which establish "effectiveness," forexample, the number of teachers trained by a teacher trainingteam during the life of a contract. These project targets arelinked in the project justifications to a sequence of higher orderobjectives, for example the number of students taught by thetrained teachers, the number of those students who found suitablejobs, their ability in later years to improve upon the performancesin the job of their predecessors, the impact of improved job per­formance on raising the welfare of the jobber's constituency, etc.The II significance ll of the project could be rated at anyone of thesestages. In principle it is preferable to rate achievements in termsof the ultimate welfare objectives, but in practice it is hopeless totry to attribute effects to causes at this level, and the evaluator isobliged to settle for intermediate objectives.

The significance of a project may have nothing to do with theimmediate effects of a project, or the effects of those effects, asspelled out in the project plan. Unanticipated indirect effects ofthe project may outweigh the direct effects. Furthermore, directeffects which were recognized but not valued at the beginning ofthe project may at a later date, and with a better appreciation ofthe development mechanism, prove to have been the most desirableof all. For these reasons significance evaluation focuses on effectsof interest to the evaluation officer, not the project officer. Theirinterests may coincide. And they may not. There is considerablevalue in seeing whether the project achieved the significant objec­tives described by the project officer, but significance evaluationis not constrained to this line of inquiry.

\

As inferred above, the "efficiency" tests can be appliedeither at the level of effectiveness or at anyone of the levels ofincreasing significance. In other words, one can refer to bothefficiency-effectiveness and efficiency-significance evaluations"At each level it is appropriate to ask whether a cheaper set of re­sources, or a quicker path, could have been used to do as muchas the project actually did.

This study omitted the compliance test. It concentrated onthe effectiveness and significance tests, and gave more weight tothe latter because they matter more. The analysis reached thepoint where it was difficult to keep the two separate for reasonsdescribed below. Nevertheless, as far as possible the distinctionwas preserved.

19

This conceptual distinction between effectiveness and signifi­cance explains why the study developed as a two-stage study, andwhy the field trips pursued two lines of investigation. On the onehand it asked about institution-building, and tried to see whetherthe U. S. extension technicians built a set of "viable" extensionservices while they were in Latin America. This is the effective­ness study, and is based on the assumption that the principal targetof most of the U. S. extension projects was to build an extensionservice that could stand on its own two feet, or put another way,that could survive phase-out of the U. S. programs. Chapter 3deals with this matter. First it lays out the issues; then it dis­cusses the findings; finally it brings to bear some evidence fromthe comparison of country experiences.

On the other hand the investigation looked at agricultural pro­ductivity, and tried to see whether the extension services the U. S.helped build have increased the level of productivity. This is thesignificance study, and is based on the assumption that the ultimateobjectives of extension services modeled after the U. S. institutionare economic and can be measured in terms of productivity. Thisis one choice for a significance test. It is explicit in some of theearly extension project proposals, implicit in some, and disre­garded in the others. It makes many of my extension friends un­easy, partly because it overlooks the leadership training, homeimprovement and other non-economic benefits which are part andparcel of the U. S. extension philosophy. Chapter 4 deals with thismattet-. It discusses what is referred to as the "impact" study:first handling information collected at the village level and theninformation collected at the district level. Next, it brings in somelessons from a large number of so-called case studies of extensionprojects. Finally, it presents tentative findings on the economicrole of extension in the study area.

Thus we look first at the institutions the Americans built,and then at their results. By using this two-stage analysis we cantie the American technicians--the inputs in the institution-buildingprocess--to farm productivity--the output of the productionprogram.

The study makes a brazen attempt at efficiency evaluation,the most difficult of the studies of technical assistance, inchapter 5. This, the final analytical chapter, is intended more tolayout the efficiency issues than reach conclusions. The analy­sis is more speculative than in preceding chapters.

20

The two-stage effectiveness / significance analysis is one ofthe fundamental characteristics of the study design. The other isthe large study area. We are mostly examining the experience intwelve countries where U. S. extension programs were operating.Multi-country analysis is essential for distinguishing factors thatmatter to the program universally from those that have influencedue to peculiar country circumstances. The twelve country sampleis perhaps larger than most multi-country analyses, but therewere significant economies of scale in the offing. Although thecomparative analysis of country experiences does not detect manydifferences, the very similarity of the results of a large numberof institution building projects enables us to draw some bold con­clusions about the overall effectiveness and significance of theU. S. extension programs in Latin America that a smaller studywould not have allowed. Conclusions drawn from a large numberof experiences usually offer a certain safety, and it is for thathaven we are headed.

Evaluation studies frequently resort to the use of the' termssuccess and failure. This study is no exception. Success pre-·sumably means that targets and objectives have been achieved;failure means the reverse. But with institutions it is impossibleto draw this line; all one can do is observe a number of allegedindicators of institutional viability and see whether one extensionservice has a greater ratio of hits to misses than another service.Similarly on the productivity issue, we find there has never beenany extension agency, or agent, that cannot claim some impact onfarm technology. The differences are in degree. Thus in a sensethe phrases "a successful extension project" and "a successfulextension service" are inappropriate. Nevertheless, they areuseful and are used. However, they are used less frequently thanoriginally anticipated. As just mentioned, the extension serviceswere generally too similar for the study's sensing devices to dis­tinguish. So success and failure are terms that appear to be morerelevant to an overall evaluation of the multi -country program,than to a discussion of individual country experiences.

C. Research Program

The first job on the research program was to identify theU. S. inputs, in particular the American technicians who visitedeach country. It required a search through both active and retiredrecord collections. It was not an easy assignment, because most

21

of the records appeared at first to have disappeared. Thankfully,people who want to destroy records are no more efficient thanthose who want to stop them from doing so, and eventually it waspossible to reconstruct the essential personnel history. The toursof duty of each technician were charted for each country. Figures1 and 2 show representative charts for Peru arid Bolivia, twocountries where the U. S. teams exhibited markedly different prop­erties: the emphasis in Peru was on subject matter expertise, theemphasis in Bolivia was on extension methods expertise (inter­country comparisons of this sort are discussed in the generalreport) .

The second step was to find and interview as many as pos­sible of the U. S. technicians. Altogether I talked with about twohundred technicians and other U. S. agricultural program personnel,three-quarters of whom are so marked on the charts (the checkednames in the two figures). Some were interviewed in Washington,some at the Missions during my two extended journeys, and a fewat their present U. S. addresses outside of Washington. Theseinterviews were structured only to the extent that the persons wereinvariably asked to describe their own work, and the work of theother technicians (as shown on the personnel charts) who were theircontemporaries in the country of assignment. The intervieweesalso were asked to comment on the successes and failures of theiroperations, and the problems encountered. Since the experienceswere usually over ten years old, the degree of objectivity andaccuracy of the responses must be assumed to have been low.Fortunately, more than one respondent was found to cover eachperiod and corroborate other testimony.

The third step was to select criteria for evaluating results ofthe institution building and production programs. The FederalExtension Service helped in this phase, as did many persons inAID and under contract with AID.

The fourth step was to select from within the study area dis­tricts suitable for-the impact study. The choice depended partlyon the availability of maps, as will be shown.

The fifth and final step in the research program comprisedthe field trips themselves. There were two, in the fall of 1968and the fall of 1969, involving altogether seven months. Theitineraries of the trips were essentially the same, and the routingis traced on map 1. The first trip was devoted mostly to the

N

os

OCE

~"""'---j-------'-O

SOUTH

-A T LAN TIC

. 'ALJlI.AIID IS•

60

.-Jr---+--------+-----05

90

latin AmericaN~RTH

.,~~~;:-:---=~!!!f~!'!-~~ A T LAN TIC(J --rt-------+-~---15

OCEAN

II 5

Map 1

.cHI ...,-...,....NOT LY~..Tn'.

STUDY AREA, DISTRICTS AND TRIPS

22

\--t-----i1\---------f---------t------30

15--------+------"G9Guatemala

/

a Sal.ad0ll-EL 5,.\.~,.Q

Sa'::1i~ci;"'~~~~~,

15 15

PA C I F IC

OCEA N

STUDY AREA 3030

• STUDY DISTRIC STRIP ROUTE

05-------+---------f---------j-}H'''I'

O-------+----GA-U-:P-AG-O-S-I-S-.~t_----__"'7,•.

57827 12~7

Figure 1

CHART Of=" 'fHE oFFIC IAL U, S. E}(TEt\JSIDfJ 'tEAM PERURo\,.>

A SCIPA o

:3

..1'I42 3 'l-

I/AA2 3 s , 7 8

c

3 s , '7 8

AItH,UIOSW SflIlLl(

~ .. 'CA-.-

f\)w

c! { .:.S.:.;...:.;lfc.::.'____-=L~v,:.::"'___ _

Ii.

~~~ .!."~ - Li"'~c7.. - - - /'o"{o A,,_

(Gll.,n) As .lO"'. tt t•

.~a!.._~~_

!!A~'~...!ci

7

10II)

1\

7

".,. .. OFIIR

'{ .(5 :_.-7'

~

Figure 2

CHARI Of 1l1E O~ICIAL U.S. €)(T"€~)lbtJ Tl:AM BOLIVIA

A SAl o

s,777<t

'610II

,c

~.~~!~ ~~~ _ <:L~.v _M~t1.-_

HU6M&'r L-K ~6L& l"Ct. ./.f"iw.;;~·..:~it : ~.::: :: _ .

.~.!!~~ !~~.; .~ ......~ -.~.~.~ E:~ .

.~.~~ ~~, .

~_{!!,.::.:.c:.:T..::....--=L,,::.:.::::k_

?10

qCO~~

oft. CD

~e

57 N10

~~

~VI

"

" 1h,oif. /

I\)W

, "7087,sIj3

o

'2.

c

~~!!!S_ ~.~.

9 1'1'0

PERU

8

c~ {S"". Lv."

Cl.

7,s'"

.........

32

~~- !."~ - "i:i., -s';;-" - - - /80Yo e,.._

CGI,..I,") A,,....Lo,... tto"'t'I

~oo~c_

BlttANT'

l~ Z'... l: ......

~

~.!!..rs~~!! _

!1.'~I!!

I'!SO

~~.v_tt.o..~~~ ..... ~~~ ..._

9

(.«....) Pa+~ \'.N'''oO. P••l, {o."~.Joih./8h..",_S,.i!!. ~_"""L~_<';,",.-.;;:c~"-.:::J.,,,-,~,,_ (,_.j "" ./'''''n _ ~._..;;:. __ ~. """,-",>-

•~ Lv,I< -- -("'''''N111)_ 0-L..." /<. ./)S....'~~D.:;:'"'L<..,:..J.,..{_ -~-_ -1iO!!!'.!!.{F!!!!.~~~~_\_((;Cou_ _ L"!-l~~"L ~_-- -rc,

SC"N...vrl . • F...~.'.t. .... "'$~Aii..-) cr~..,t~....,• .:....<--... ~\",; ..... ~~ "G..-':t'../·",····,·;~~···o;" ~. "_I{• ,.." CA"""N'- -~ ­

{,)N,"I!~{,. ,i{ """~.-Z{j~' • t/ "{ ~",,,.;jLV"&<-Y"'L1·1l '~~'".", _ ~ ~..~ ...~ l!'!E __ '")

"'{ r)A.,..-!/

I(IIA' " "_;.. ...

! ~...JL·tl::k..- . ,..... .,,~'&llu~ ",A1f'J 1'........ 6>"'''''1"""'--~, j ~.r (.-'l..... •"""~ ---- - , '- .~ G ...y Lv...._.- -"""- - - - ~. .' '_","!!!v=.. ;.;;;.,,__

- - ;:, {C,._) Lv,..

<t ...

~ ~

8

Awoet«~'"E"t-- ---

7

.. ~o!._~':..._

~,~~~

ffUsreL&,

,

~/

~

CHARI or- 'fHE OFFICIAL U.S. El<TENSI/:>tJ -rEAM

Figure 1

SCIPA

v n''''lE

B fll,.+e""".", Sot'vlc.e

'"_",,,Il Oid~

GVf'loI'U£

fll(W.ROSe-t U"oc UAAItK"'" f',n"Y.f:~!~.,! .. Clb'o-.- --

!!!!.~:::.l~'~'~'~~"""''''''''''''''''''''''''''''''''''''''''''''''''''''''7'·..·...... ·....· ........ Uti .~:~¥'.i~•.~~.,L"" u~ / Roo••n ./ •GA." P'KS ;;OOU"IV/ ; "(un' _ "III",SI.'- 1'1e••12 '"'cc O-IOJ \/.............. ..... w.:..." r. ..... >t... GUf.OMeL

en_I_) (lW'Nr<)./ ("IL'.") UtI<C) .<~.a..,.lI)4.........o_) ./

'- (;UfO&R"r2. h , '(ODS'S."') .1)(. ,.. " tLA(.UN" "'0"'. SC.Mt¥u1'J. "'_ ~."""" (UZ'ZL5) '4.~..

,/

00-11I'-" .......SUNO" -- (,4<.......0) 1.,(.

--.. '-. \ /J~."!!! J!~ .

____'" 6"'.0') s.....

\

A

'" ~l 3 ~

IIAA

It'

~...

7

1010II

'7/I

.,.. OFAR

'{5 : ,..... /

7.

\:

ROI»

'2.

S5,b

S

:3

"

25

Figures 1 and 2

Notes

Row Detail:

Row 123

5

67

1011

Length of life and name of the agricultural servicio.Length of life of the servicio extension service.The interval between the dates refers to thecalendar period of the earlier year.Agriculture team leaders. During the servicioera, these men were also directors of the servicio.Their deputies.Extension advisors, including area program advisors.Home economics advisors.Agricultural information advisors.

Unnumbered Rows:

All other persons are subject matter specialists ortechnicians working in related fields. Abbreviatedtitles are shown.

Underlining

Row 6

Rows 7 -11 plusthe unnumberedrows:

Principal officers who concentrated onthe extension program

Over 70% of time devoted to extension

Between 30 and 70% devoted to extension

Up to 30% devoted to extension

Others Signals

./ Interviewed for study.

( ) Contract, consultant or post 1960 PASA personnel.All others are AID direct hire.

Sources: Active and retired U.S. Government records, and inter­views with U. S. and host government personnel. Printedstaffing patterns are available in Washington for at leastone month each year since 1952.

26

institution building issues. I interviewed widely in each of the hostcountry capital cities, using mostly an unstructured format, butreferring to my charts and other literature when .the host officials Iimpressions of individual U. S. technicians, servicios and otherphenomena were needed. The second trip, the longer one, wasdevoted mostly to the impact issues. The seven village levelstudies were made on this trip, as well as most of the case studies.However, additional material describing the state of health ofnational extension services was also collected in a highly structuredfashion. Appendix B contains the form that was completed in everyextension headquarters, and the questionnaire that was passed out.I have never counted the number of persons interviewed in the six­teen countries visited during the two trips. The figure is probablyabove three hundred, the majority of whom would be nationals.

Large amounts of documents and other paper work wereassembled during the field trips, and sent to Washington for useduring the period of analysis and write-up. In spite of this substan­tial library of papers and interview notes, there was no real re­lief from the data problem which characterized the entire study.

I collected every piece of information having to do with ex­tension, from diverse sources that differ from country to country,and tried to put them together in ways that are revealing. In somecountries I travelled with extension agents. In others with fertili­zer dealers. In others with credit supervisors. In some countriesI talked with farmers, in others with shopkeepers, in others withneither. Companions and the interviewees were chosen in somecases because they offered the most relevant experience, in othercases because they were all that were easily approachable. Insome countries I found the old extension service1s daily log ofyisits to communities. In other countries I couldn It, so I had tofind the old extension agents and ask them to remember where theyhad been. This sort of study, in short, requires a first order jobof scrounging. To use the word methodology is misleading.

D. Study Area

All countries in Central and South America are treated inthis paper with the exception of the three Guianas and BritishHonduras. The island republics, in particular Cuba, Haiti andthe Dominican Republic, are excluded. With these few omissions,the study covers Latin America.

27

Attention is given to the string of twelve, contiguous Spanishspeaking republics running along the Andes from Guatemala toChile. Nowhere else in the world can one find a comparablenumber of independent nations which share so many geographic andhistoric characteristics. It is not a homogeneous group, of course,Colombia, the most populous, had twenty million people in 1969and Panama, its neighbor, had one and one-half million. CostaRica is mostly white- -that is, of unmixed European ancestry.Guatemala, Ecuador, Peru, and Bolivia are at least half Indian.Paraguay does not extend to the Andes, Bolivia does not extend tothe sea, and many of the people of these twelve countries live eightthousand feet higher than the others and eat potato instead of corn.Moreover, it is dishonest to call the entire chain of mountains theAndes, since, strictly speaking, the Andes run north along thewestern coast of South America only as far as Colombia, and thecontinuation of this .geologic phenomenon through the Isthmus andCentral America goes by other names.

Nevertheless, the characteristics these countries share,especially the economies and cultures of the settled or detribalizedrural areas, are striking. Extension supervisors, extension agents,the receptive small and middle sized farmers they gravitate toward,and the corn and potato fields these farmers cultivate, are indis­tinguishable from country to country. Chile, with higher incomelevels and few subsistence farmers, is rather different and wouldhave been dropped from the group except that its experience withU. S. extension advisors resembles the others, and provideslessons.

The twelve countries - -Guatemala, El Salvador, Honduras,Nicaragua, Costa Rica, Panama, Colombia, Ecuador, Peru,Bolivia, Chile and. Paraguay- -contain 72 million of the 249 millionpeople in Central and South America in 1969. The twelve are re­ferred to in the paper as the II study area." They are treatedequally, which may be a mistake since the lessons from Colombiadeserve greater weight in the overall appraisal of the program thanthe lessons from individual, small Republics. The reader shouldbear in mind that such a weighting process was not performed.The study area is outlined in map 1.

The other 177 million people live in countries whose ex­posure to U. S. extension programs was significantly different-­usually lower--and which are included in the study for purposesof comparison with the principal twelve. Mexico, Venezuela and

28

Argentina are handled together. They never received officialU. S. extension advisory assistance or U. S. financial support forthe extension budget. American advisors in private capacity orattached to private foundations were active in all three, though ata lower level than the official U. S. teams in the study area. Thesethree countries have extension services, and can be considered"control cases" because they help point up the advantages and dis­advantages of official U. S. assistance. The comparative analysisis less interesting than it might have been because the threecountries are among the five wealthiest in Latin America in percapita income and government revenue, factors critical to theviability of extension services.

Brazil, with 91 million people, has a flourishing extensionprogram, though the extension service is under state control andthe quality varies from one to another state. Because of theauthor's inadequate control of the language, the confederal charac­ter of extension administration and the sheer enormity of the task,the Brazilian extension experience has been ignored here. Notentirely ignored. It appears that several of the Brazilian stateservices are superior to each one of the national services in thestudy area, and it seemed useful to try to specify the reasons whythis might be so.

CHAPTER 2

U. S. PROGRAMS IN EXTENSION

A. 1940-1950

The first permanent teams of official, U. S. agriculturaltechnicians working in foreign aid programs arrived in LatinAmerica in late 1942. They were preceded by several reconnais­sance missions, which discussed with interested local governmentsthe possible assignments and locations of the experts. Twoagencies of the U. S. Government were involved at that time: theU. S. Department of Agriculture (USDA) and the Institute for Inter­American Affairs (llAA).

The USDA organized its hemispheric aid programs underthree titles. Most important was the program of the Office ofForeign Agricultural Relations (OFAR) --the predecessor of today'sForeign Agricultural Service. OFAR teams ranged up to ten pro­fessionals and engaged primarily in the development of, on the onehand, coffee, cacao and other edible, tropical tree crops, and, onthe other hand, chinchona (quinine), insecticide, hemp and otherspecial purpose crops, none of which could be grown in NorthAmerica. These efforts were collectively called the "complemen­tary crops" program. OFAR established itself in strength in tenLatin American countries between 1943 and 1948. Its trademarkwas the experimental station, ofttimes located in areas remote tothe existing lines of communication, such as "Pichilinque" inEcuador and "Tingo Maria" in Peru. OFAR teams at first stressedthe development of a biological research capacity. They limitedextension work to the dissemination of their discoveries amongfarmers in the vicinity of the stations. The emphasis later beganto shift, and by the end of the 1940s OFAR had begun in fourcountries to get involved with extension at the national level.

Second, there was the program of the USDA's Bureau ofPlant Industry (BPI). BPI teams, smaller in size than OFAR's,but active initially in more countries, concentrated exclusivelyon rubber. They also built, or converted sequestered haciendasfor, experimental research stations. To a greater degree thanOFAR, BPI involved itself in the organized extension of rubbertechnology to small and large scale farmers living on land thoughtsuitable for this new crop.

29

30

Third, there was an ad hoc program for sending occasionalteams of U. S. experts in response to special requests from LatinAmerica governments. In the study area, the only importantinstances of such activity were the teams sent to Colombia, one in1944-45, another in 1947-48, and another in 1950.

The lIAA had been chartered under authority granted by Con­gress in 1942 to the Office of the Coordinator of Inter-AmericanAffairs. It was a public entity, incorporated in Delaware, underthe direction of the Coordinator, Nelson Rockefeller. In 1946, whenthe Coordinator's Office was closed, lIAA was shifted to the controlof the Department of· State . The two objectives of this glamorousinstitution were to assure medical and food supplies for Americanand host country war workers engaged in Latin America, and topromote the welfare of the U.S. 's sister republics through coopera­tion in technical progress. It specialized in three sectors - health,food and education. The Health and Sanitation Division was earliestoff the ground, and throughout the war period was the largest opera­tion. The Food Supply Division's first overseas resident missionsarrived in Costa Rica, Paraguay, Brazil and Honduras in late 1942.The Peruvian mission, destined for fame, was established in May1943, ten months after OFAR's own Peruvian team was fielded.

The lIAA agricultural programs differed strikingly from thoseof OFAR. They emphasized production of local food crops, and thecreation of rural agricultural services which would institutionalizethe technical transformation of the Latin American economy. U. S.self interest was written over this program too, since the SecondWorld War was on, and secure sources of basic foods for rubberworkers in the jungle, for urban centers, and for U. S. garrisons,were considered essential to U. S. interests. But there was moreof a Messianic spirit to IIAA's hemispheric program than to OFAR 's,and, especially after the war, its technicians were closely associatedwith the deve lopment of the national agricultural economy.

During the war, lIAA placed agricultural teams in elevencountries, but in the late 1940s budget cutbacks forced it to retrenchand confine operations to Haiti, Costa Rica, Peru, and Paraguay.The trademark of the IIAA came to be the cooperative "servicio", abilateral, operational agency effectively under American directionwith headquarters in the capital city, a network of field offices, anda number of exotic operations in different provinces. Extensionadvisers played a dominant role in that structure. The first agricul­tural servicio was set up in Paraguay in December 1942.

)31

B. 1950-1958

In 1950, OFAR, BPI and IIAA operations came under theauthority of the newly established Technical Cooperation Adminis­tration. As can be imagined, a "friendly" rivalry had developedbetween OFAR and the Food Supply Division of IIAA. It was moreevident in Washington than in the field, since there was only onecountry where two teams coexisted for more than a few years (Peru).After President Truman's Point IV address in January 1949, whichopened the way for a major expansion of the U. S. technical assist­ance program, the USDA and IIAA struggled for control of theagricultural empire in Latin America. IIAA won. The author isnot familiar with that battle, except to say that it took three or fouryears in some countries where IIAA had thitherto not operated forOFAR management to be replaced and OFAR technicians eitherforced into a "foreign service" category (under which IIAA officerswere appointed) or out altogether. Some seasoned OFAR personnelpreferred to be repatriated rather than to sever ties with USDA andthe U. S. professional agricultural community. They did not lookforward to building a new career among the "amateur" agriculturalistsof IIAA.

IIAA was itself in transition, however. It lost its effectiveautonomy in 1951. It continued, with.the same principal staff, asthe Latin American arm of TCA and in the minds of men throughoutthe 1950s as an elitist wing of the American global technical assist­ance program. But IIAA field units succumbed in the early part ofthat decade to the formal structure of resident country missions. Agood reference point for this institutional development is 1953, whenthe new Eisenhower Government combined TCA and the Mutual SecurityAgency (MSA) to form the Foreign Operations Administration (FOA),and universalized the country mission and programming conceptswhich had grown up with the Marshall Plan and successor agencies,including the MBA.

If USDA units coexisted in the early 1950s in the same countrywith IIAA, their programs were integrated into the servicio. Thelatter became the standard administrative model for U. S. agriculturalassistance throughout the hemisphere (but not elsewhere~). It differedfrom the old servicio model in that its director reported to the U. S.country mission instead of the Food Supply Division in Washington.However it retained considerable independence over the dispositionof the joint U. S. - host government budget, and in this sense was stillfree of many U. S. legislative and bureaucratic procedures.

32

The serVlClOS survived both the creation of FaA and the re­organization two years later when ICA was formed. However, theyin turn succumbed in the late 1950s to mounting pressure on U. S.overseas missions to shift from operational to advisory assignments.By the time AID superceded ICA in 1961, U. S. participation inservicios had for the most part ended and U. S. agricultural teamshad become an integral part of the central overseas missions andsubject to the priorities imposed by the rapidly expanding staff ofprogram officers in mission headquarters.

Figure 3 shows the historical relationships between the severalU. S. Government agencies administering technical assistance to theagricultural sector in Latin America. The figure indicates the periodsin which each of these were operative in the countries of the studyarea. It also shows the length of life of the twelve agricultural serviciosset up in that area. More will be said about servicios, since they arethe center of the controversy over the alleged failure of U.S. institu­tion-building activities in the 1940s and 1950s.

Shifts in administrative organization were paralleled by shiftsin the emphasis of U. S. agricultural programs. The extension com­ponent played an increasingly important role after 1950, especiallyin countries where previously only aFAR and BPI had located. By1958, permanent U. S. extension personnel were established in everycontinental Latin American country except Mexico, Venezuela,Argentina and Uruguay, the four cOlIDtries where U. S. Governmentagricultural missions had either long since withdrawn or lived onlyan occasional and temporary existence. These U.S. technicians wereinvolved in building and strengthening agricultural extension services.They usually operated at the national level, though in several of thecountries with larger territories or populations, they concentratedduring certain periods on specific I1demonstration l1 districts. Theincreasing stature of the extension team within the U. S. agriculturalprogram reflected the victory of IIAA headquarters over aFAR, andthe consequent increase in the influence of those officers who supportedrural development programs based on farmer services, and corre­sponding decrease in the influence of those who stressed the importanceof the development of new technology. Figure 4 and table 1 try toillustrate the magnitude of this shift. They show the number of man­years of U. S. extension personnel in the study area during the thirtyyears of this program. These include both technicians who were de­fined as extension advisors in the job descriptions, and those otherU. S. personnel, who were not described as extension advisors, but

.. AID

33Figure 3

u.s. AGRICULTURAL ASSISTANCE AGENCIES AND SERVICIOSTwelve Country Study Area

1970I

u.S. Assistance Agencies

1}J?P.~l.~:r.l..(~~l:>J?~~__ f'F.<?g~.l~,I!lJ. ..

_lLSJ2.A[OtAB.. - - - - - - - - ....>....;...d(I~IAA!:£f~)===(::!:Bur=~ea==u~f:::o~r~La=t!::::i~n~Am~e=r=l='c=a=)==11M ~

.. TCA lFOA .. ICA

Country Agricultural Programs and Servicios

STICA PARAGUAY

..······..···..........·....····..····si'IcA··········.

~ _ .-- scrPA- - - -- - -'.... PERU--'~iIiiiil... - . _. _. _. ---t!

(SIPA)COSTA RICA

SAl '.- - _Ill'--.... ~••••;;.. ..JB9~L~W&~~

STICA HONDURASiiiii . _. _. _.

_OtA!Vkr.¥.!.1S~1L_ __ (DESARRURAL). SililiiCAIiilIiIP__".... .PANAMA

CHILE

ECUADOR

NICARAGUA

DTICA

...·....·..··sw·..·····

. - - - - - - - - - - --...SiiiCIA.... _

STACA COLOMBIA

- - - - - - - - - - - - - ~.;;.iII·SliCliilIiiDAiIIII__

~-------------,S~SA

.................._ - .

I1970

, I

1965

GUATEMAIA

EL SALVADOR

I I1960

I I1955

I I

1950I I

1945I I

1940

34

Technicians (Man-years)

.\Total(Direct Hire,

PASA, Contract,Consultant)

196019551950

u.s. TECHNICIANS WORKING ON EXTENSION IN THE STUDY AREA

Figure 4

20

40

30

50

60

10

Source: Table 1

35

TABLE 1

OFFICIAL U.S. AGRICULTURAL TECHNICIANS IN THE STUDY AREAShowing the extension team as a percentage of the total.

Agricultural Technicians Extension Team Technicians

YearDirect Hire . Contract Total Number+ Percentage of

USDA· IlAA FOA-ICA-AID Total University Other TotalTotal

(I) (2) (3) (4) (5) (6) (7) TBT -m- (IO)

1942 8 8 843 5 15 20 20 1 5.044 20 60 80 80 4 1/2 5.6

1945 29 45 74 74 4 1/2 6.146 28 25 53 53 6 11.347 25 25 50 50 8 1/2 17.048 35 22 57 57 13 1/2 23.749 33 31 64 64 15 23.4

1950 35 33 68 68 16 1/2 24.351 34 32 66 66 19 1/2 29.552 64 54 118 8 8 126 39 1/2 31.453 75 54 129 16 16 145 44 1/2 30.754 158 158 14 14 172 51 29.7

1955 146 146 14 2 16 162 49 1/2 30.656 180 180 24 11 35 215 61 28.457 204 204 29 15 44 248 64 25.858 214 214 21 18 39 253 59 1/2 23.559 209 209 13 26 39 248 56 22.6

1960 210 210 5 29 34 244 55 22.561 179 179 4 32 36 215 55 1/2 25.8 -62 176 176 6 28 34 210 42 20.063 148 148 5 29 34 182 34 18.764 3 110 113 26 16 42 155 30 1/2 19.7

1965 11 72 83 43 28 71 154 24 1/2 15.966 18 56 74 51 28 79 153 19 12.4

Note: Figures in columns 1-8 show number of persons on board on January 1. Figures in column 9 referto July 1. Column 8 excludes consultants and local contract personnel; column 9 does not. Since thereare a few of these persons, the two columns are (slightly) incomparable.

+ Technicians devoting betweenconethird and two-thirds of their time to extension activities arerated 1/2; if more than two-thirds they are rated 1; if less than one-third they are rated O.

• Excludes USDA Bureau of Plant Industry rubber technicians before 1954.

36

who worked part or full time, either alone or with the extensionteam, on extension type activities. The reader will note thatbetween the years 1950 and 1958 the annual level of extensionadvisory assistance rose from 20 to 60 man years. The weight ofthese extension activities in the total U. 8. agricultural programincreased from ten to thirty percent.

The curve of figure 4, illustrating the conscious set of deci­sions starting in 1950 to promote extension, and the equally consciousset of decisions around 1958 to get out of it, is a one line summaryof the subject of this study. It is that pyramid of men and women thatheralded the "golden years" of extension in Latin America, a phrasestill used somewhat sadly by survivors in the present Latin Americanbureaucracies, many of whom have never understood why U. 8. interestsand their own fortunes could have collapsed so quickly.

Characteristic, and not an unimportant explanation, of the risinginfluence of extension in the early 1950s was Mosher's study, preparedin 1953 and 1954, on U .8. technical assistance to Latin America, inparticular the pair of chapters describing the OFAR program inEcuador and the lIAA program in Peru. 6 The first he compares un­favorably with the second, largely because the latter was allegedlystructured to move information to the farmers while the former in­adequately and rather informally acquitted this essential task. Mosher'spub lished criticisms of the Ecuador program were in fact anticipated.Within one year of his first study visit to Ecuador, the officer incharge of the program was replaced by one of the most dynamic (aswell as controversial) individuals in the history of our foreign aidprogram, a man who, though not a trained extension agent, gave theU . 8. program in Ecuador a strong field bias and drove all his U. 8.technicians and their Ecuadorian counterparts into the countryside tomake contact with the farmers and develop with them profitable pro­jects which made use of the best of whatever technology was thenavailable. What Mosher could not have predicted was that in somecountries, but to a less extent in Ecuador, the extension bias would bepushed too far, getting as far ahead of the research base as t~e latterhad supposedly once been ahead of it.

6Arthur T. Mosher, Technical Co-operation in Latin-AmericanAgriculture (Chicago: University of Chicago Press, 1957), pp. 39-72,181-190.

37

C. 1958-1970

After 1958, the influence of the extens ion philosophy on U. S.assistance strategy waned. Critics pointed behind the official reportsto continuing sluggish agricultural performance and underpoweredancillary government rural development programs. Whatever theconceptual merits of the extension program, it had not demonstrablyaltered the level of productivity or the structure of the rural economy,and it had not persuaded the host governments to give any more atten­tion to the traditionally impoverished rural sector than they had inthe past.

The change in attitude reflected also a restlessness within theU. S. professional community with respect to the unchallenged asser­tion of professional extensionists about the role of extension in U. S.domestic progress. The end of the 1950s introduced what one authordescribes as "the period of great debate concerning extension's roleand functions." Impatient for res ults, the U. S. aid agency (ICA)sought quicker solutions to rural stagnation. As shown in figure 4most of the extension technicians were withdrawn over the seven-yearperiod ending in 1965. The timing of this withdrawal was not identicalin all twelve countries. In fact in Chile the U. S. extension team didn'tstart work until December, 1957. But the writing was everywhere onthe wall. By 1965, the number of extension advisors had been reducedto ten, the 1947 level. Two years later there were only five occupiedextension positions listed in the staffing patterns of the twelve countrymissions. Not all the extension advisors were repatriated, thoughmost of them were. Some were reassigned to other jobs in the countryteam. Extension projects which had been labeled as such were droppedfrom the mission reports. In what must be the greatest disappearanceact in the history of AID and predecessor agencies, a team of eightU. S. extension advisors to the national extension service of the Ministryof Agriculture in Bogota, Colombia vanished from the record of annualprograms submitted by the USAID mission between 1962 and 1963.

This paper is not the place to launch an attack on those aspectsof the AID programming process which made possible the frequent,repetitive, and often devastating discontinuities in U. S. technicalassistance programs. These discontinuities -- in leadership, techni­cal personnel, organization, direction - - are only apparent when therecord of ten or twenty years is examined. They were apparent inthis study of U. S. agricultural programs in Latin America, and pro­bably will appear in a study of any sector program on any continent.

38

They are not apparent to most AID field officers: since those officersare not in the habit of tracing the history of projects more than a fewyears; since the Agency's record retention system does not easilypermit them to do so if they wanted to; and since most officers remainat a job for a period of years even shorter than the already briefcycle of a project life. Why this happens, and what should be doneabout it, are subjects for other reports. It is important to note thephenomenon here, however, since it was particularly painful in itseffects on the extension program.

The two main inflections in the curve in figure 4, in 1950 whenTCA got into extension and in 1958 when ICA started to get out, arethus the results of two independent forces. One is the rational calcu­lation of development needs and strategies, with an evaluation of theproper role and performance of extension services within those stra­tegies. The other is the eagerness of some AID programmers toimprove upon programs and projects inherited from predecessorswithout recognizing that uncertainty and failure are endemic in thedevelopment profession and that their special visions, sometimesvisible only to themselves, more often than not turn out to have beenno better than the visions of earlier men.

Analysis of the period starting in 1958 is complicated by thefact that several other elements of the previous rural developmentstrategy, indeed of the Agency's overall approach to foreign aid, wereunder attack. Extension activities were not the only part of the pro­gram with a shortened future. The servicio type of operation, indepen­dent of both U. S. and local government standards and controls, waswidely criticized as the perfect way not to build a capacity within thelocal government for designing and administering programs. a capacitythat could survive U.S. phaseout. The argument over servicios wasbitter, since most Americans who had worked in one were staunchdefenders of the system, and most who had not were quick to criticize.It went on for several years, ending finally in a blanket order that oneof the factions in Washington was able to get out instructing the missionsto terminate servicio operations as soon as possible.

The concept of foreign aid as primarily the transfer of technicalskills and institutions was also under review. This concept had neverbeen part of the Marshall Plan experience, and U. S. aid programs inthe rest of the developing world had generally mixed technical withcapital aid. There was no important capital program in Latin Americabefore 1958, however, other than for "emergencies", and many inWashington were wondering if it was not time to abort the traditional

39

IIAA approach and move into the big leagues. This meant infra­structure projects, country-programming and economic analysis.It meant the end of large, rural field staffs, rural based projects,and strategies designed as well as run by technicians.

Finally, by 1961 there was a new, Democratic, governmentin Washington and a major new initiative taking shape under the pro­posed Alliance for Progress. ICA was in the process of reorganiza­tion. The youthful administration was eager to seize the opportunity.It wanted to replace what appeared to it to be the rather ineffectualLatin American program of the previous decade with something better.Rather than try to build upon the strengths of the previous program,it pressed for a bold departure.

The extension program, formerly the centerpiece of the Agencyfsrural development programs and, therefore, probably the strongestcomponent in the entire U. S. technical ass istance effort in the hemi­sphere, did not survive. I have been advised by those who were closeto the decisions that extension was beaten not only by its enemies butalso by irreversible forces within the Agency searching for new de­vices to promote development.

In one country - - Paraguay - - the Ifold If type of U. S. extensionproject continued. The servicio survived there until January 1969;U. S. budget support to servicio operations, including the country'sonly extension service, was maintained (though without new dollarobligations after 1967); and at least one U. S. extension advisor hasbeen at post. In four countries - - Guatemala, Ecuador, Bolivia andHonduras - - the U. S. continued to provide budget support for theextension service for several years after the termination of theservicio. The argument in Bolivia was that the continuing failure ofthe government to reduce the central government deficit made it likelythat without U.S. budget support the extension service (which had beentransferred from the servicio to the Ministry of Agriculture in 1961)might not survive.

The experience in Peru was unique. There, the contract withNorth Carolina State University was rewritten in 1962 to allow thatuniversity to support'the local extension service. Thus by 1963,when U.S. involvement in extension in the rest of the study area wasabout dead, North Carolina extension personnel were arriving inPeru to initiate a joint program for rural development in the southernsierras.

40

North Carolina's initiative with the extension service in thesouthern highlands of Peru itself died in 1965. But we can think ofit as the beginning of a new set of relationships between the U. S.government and the extension services of Latin America.

In the last five years, AID has reestablished a presence inthe extension community. The form that presence takes varies fromcountry to country. It invariably has started at the margin, or inspecial projects, rather than in direct and substantial aid to strengthenthe central extension service. And it has called on the USDA and theU. S. land grant universities to supply extension personnel to supervisethese new programs, rather than build up an in-house staff of AIDemployees as was formerly the custom. The new projects are oftenexperimental. In Guatemala, two experts from California Polytechnicare helping to develop a mobile extension school to train farmers inthe eastern provinces. In EI Salvador, a former resident AID agri­cultural officer was called back to help set up and supervise a nationalcrop campaign organized under the aegis of the extension service. Acontract technician from the University of Florida was the primemover of the 1967 corn campaign in Costa Rica, also run by the localextension service. A group of experts from Utah State University havebeen working with certain units of .the Bolivian extension service inpopularizing new instruments and markets for the wool industry, onthe one hand, and new, improved wheat seeds, on the other. In threecountries, EI Salvador, Nicaragua and Paraguay, AID is now fundingadvisors to the director of the national extension service. In eachcase the advisor is supplied by the USDA under a Participating AgencyService Agreement (PASA).

All these examples of AID's present support for extension inLatin America are indicative of the change that has occurred since1960. For a period of six or eight years, the U. S. Government hasavoided direct and massive technical and budget support. To someextent the slack has been taken up by the Food and Agriculture Organi­zation (FAa), which has placed top level extension advisors with abouthalf of the twelve countries in the study area, and by the Inter-AmericanDevelopment Bank (BID), which has provided financial resources forexpanding and equipping extension services throughout the hemisphere.But now it is apparent that the Latin American Bureau of AID intendsto assume a prominent role in helping redesign and finance ruraldevelopment programs. In four countries of Central America alone,major AID loans have been made to underwrite an agricultural sectorstrategy in which extension is intended to playa decisive part. AGuatemala loan of $23 million for such purposes was signed in 1969.

41

A Costa Rica loan of $16 million was signed in 1970. The form ofthe new U. S. presence in extension activities differs from the oldin many respects. including the de-emphasis of advisory assistance.and the new strings tying whatever technical assistance is providedto the explicit goals of the sectoral strategy. Nevertheless it mustbring some comfort to American and Latin survivors from the goldenyears to see the U. S. government tooling up again to promote exten­sion in the Andes - - albeit in new ways and under new direction -­and in a few countries rescue a faltering institution which the U. S.had initially sponsored two decades before.

D. Financial Assistance

In addition to providing technical assistance. the U. S. govern­ment made a substantial financial commitment to the extension services.The bulk of these monies was transferred under the servicio arrange­ment. The typical servicio. whether in agriculture or not. was fundedjointly by the U. S. and host governments through special appropriations.It was common practice for the U. S. to assume the major financialburden in the early years of a servicio. ranging up to 100% of the totalbudget. Thereafter the U. S. insisted that the host government graduallyincrease its contributions and share. By 1958. the U.S. share haddropped nearly to zero in a few countries. despite the fact that theservicio budgets had grown in absolute size. Where the extension ser­vice was organized within the servicio. its budget was financed from theordinary servicio pool. There was no earmarking of U. S. monies forspecific servicio projects. Thus the U. S. contribution to the extensionbudget must be assumed to equal the fraction of that budget correspond­ing to the share of the U. S. contribution to the overall pool. Table 6on page98 shows cumulative estimates for each of the twelve countries.In a few countries. the U. S. was unable to reduce substantially itscontribution without running the risk of killing the servicio program.When in those countries the extension service was transferred to thehost government. or when the servicio was closed. if that happenedfirst. the U. S. continued to support the extension budget via directdollar grants to the host government. These post servicio contributionsare also estimated in the table. Bolivia is the worst "hardship" case.though Paraguay could be included in that group since U. S. local cur­rency contributions still continue.

42

I have closed the cumulative fiscal series for each country inthe table so as to avoid including the new forms of capital assist­ance for extension services which AID has experimented with in thelate 1960s. These would include theshare of counterpart fundswhich have been allocated by the Colombian Government to the newInstituto Colombiano Agropecuario (ICA) extension service, andthe funds earmarked for extension in the new sector loans toGuatemala, Costa Rica. Panama. etc. This still leaves severalearlier extraordinary capital transfers that occurred during the1950s outside of the servicio mechanism. Among these are theemergency relief program in Bolivia (1954 -1960) which in partbenefited the Bolivian extension service, the post-Arbenz programin Guatemala. and a few others. They are shown as separate en­tries in the table.

The extension service in EI Salvador and Panama were notpart of the servicio in those countries, and received very littlebudget support from the U. S. government. In Chile the extensionagencies within the geographic area covered by Plan Chillan wereencompassed in that Plan and enjoyed material support from aspecial fund in DTICA, the Chilean agricultural servicio. It is notproper to liken this to budget support in the other nine countries.In those nine, the weight of the U. S. contribution varies. So do thenumber of years during which the U. S. government, either withinthe servicio mechanism or outside of it, continued the contribution.It is therefore impossible to make meaningful comparison of thecumulative U. S. contributions. The total twenty-five year, twelvecountry, capital costs to the U. S. were $15 million.

CHAPTER 3

EFFECTNENESS: BUILDING EXTENSIONINSTITUTIONS

A. Appropriateness of the Study

The principal objective of the U. S. extension program inLatin America was to organize extension services that could com­municate with individual farmers, and thus close the gap betweensources and users of new technology. The services were organized,typically, outside the regular government bureaucracy where itwas felt the initiative would be buried under red tape. The empha­sis was on getting something of value going, quickly, as it was forall servicio projects.

The Americans who formulated the extension program, andthe servicio structure itself, were accused later of having sacri­ficed enduring institutional improvements for short-run productiongoals. The accusation is unjust on two counts, since it implies in­correctly that an emphasis on the short-run necessarily precludeslong-term results, and that the Americans were not interested inleaving anything behind. They were, and in fact some of themargued that the best way to guarantee a permanent place for exten­sion in government's institutional structure would be to demonstratethe role of extension in an environment where the factors makingfor failure would be minimized, which to some of these personsmeant outside the ministry, away from the influence of politicians,and under the direction of skilled foreign technicians and selectedlocal leadership. In retrospect, it appears that they were wrongto assume that government bureaucracies respond to successfulpilot projects, 7 that the transfer of institutions from the serviciowould be smooth, and that the American presence would continuefor several decades. But this set of errors is different from the"short-run" allegation.

7Robert J. Shafer remarked "Changes in the administrativehabits of Latin Americans come not as they are surprised by learn­ing of other methods than their own (they have heard of all methods),but as they feel driven to adopt something different by pressuresgenerated in their own environment. "__-The Servicio Experience(Syracuse: Syracuse University, Maxwell Graduate School, June1965), p-,-~_7._

43

44

When the servicios were denounced for their inability todevelop permanent institutions. a better line for their detractorswould have been to suggest that even though the servicio managershad created viable organizations. those organizations were onlyviable in the servicio context and could not survive project trans-fer on servicio phaseout. This would have set the debate up properly.as a dispute over alternate strategies to build institutions.

I have concluded that the early American extension teamswere oriented at least by implication toward institution building.Moveover. during the hey-day of extension in the mid-1950s. theU. S. programs clearly stressed this aspect. Thus. institutionbuilding criteria are the appropriate tests of program effective­ness.

B. Extension Organizations

In all twelve countries. national extension services are nowin operation. units within the public sector that are autonomous atleast to the extent of planning their own activities. Except forColombia and Chile. these services are lineal descendents of theextension organizations that were built with official American help.There was nothing resembling the standard extens ion service modelprior to the American arrival. and the current organizations arethe result of American intervention. While extens ion serviceswere probably due to emerge one way or another. they would cer­tainly not have started as early or developed the way they did with­out that intervention.

Figure 5 shows the growth in the number of national extensionservice agencies in ten of the twelve countries since the firstagency was opened in Peru in 1943. The figure reproduces thetrend in the number of U. S. extension technicians. because a compar­ison of the two trends indicates that the withdrawal of U. S. per­sonnel and financial support which began in the late 1950s had nopronounced effect on the growth trend in the number of agencies.These figures are aggregates. and they hide the temporary butsevere reductions in agencies that occurred in Bolivia. Colombiaand Honduras. They are also misleading in two respects. First.they ignore the other effect of the budget cuts--the losses in salaries.operating expenses and vehicle replacements. The latter were de­cisive. of course. and it can be argued that the continued expansionof the number of agencies in the face of budgetary difficulties was a

Source: Miscellaneous material

U.S. Extension Technicians(Man-years)

U. S. Technicians(All Twelve Countries)

197019601955

Figure 5

.p',.ri.:

.: ')ii

if:

NUMBER OF EXTENSION AGENCIES IN THE STUDY AREAShowing also the number of U.S. extension technicians

1950

40

60300

100 20

200

400

Agencies

46

mistake. Second, they are inclusive of the effects of latter dayforeign aid, particularly the BID money which began to arrive (inPeru) in 1966 and, to a lesser extent, the new AID agriculturesector support, both of which were tied to the creation of newagencies. A chart of the number of self-financed agencies in thepost-1960 period would undoubtedly show less growth than thatindicated in figure 5. Yet it does seem proper to conclude that atleast part of the argument about the importance of the extensionpresence had taken hold in Latin America, and that, after a year ortwo spent adjusting to the end of the bilateral program, local govern­ments were willing to protect and even expand that presence. Afew of the salient characteristics of the individual extension servicesare summarized in table 2. They refer to 1969.

C. Application of the Criteria for Success

AID can take credit for having helped support the most impor­tant theoretical work that has been done in the field of institutionbuilding in developing countries. A study of the process of institu­tion building - - of its characteristics and of ways to measure itssuccess -- was commissioned by the Ford Foundation in 1963, in acontract with a consortium of four universities organized as theInter-University Research Program in Institution Building underthe initial direction of Milton J. Esman at the University of Pitts­burgh. Esman and his associates developed a conceptual frame­work for the institution-building process (referred to as the IBmodel) and identified those variables thought to be decisive in de­termining institutional strength. In 1964, AID financed case studiesof four institutions built with U. S. support in four countries, studiesundertaken by the consortium and based on the IB model.

I will not describe the model here, but only point out some ofthe general issues raised in the IB analysis. 8 Most of the impor­tant IB variables will be separately discussed in succeedingparagraphs.

8An AID/Washington circular Airgram in the Program Evalua­tion series (PE #7) highlights these and other issues in a messageto USAID Missions intended to stimulate adoption of improved evalua­tion techniques (Robert L. Hubbell, "Program Evaluation-InstitutionBuilding" [AID CIRC. XA-4247, Aug. 31, 1968]).

* Included in row (10)** Argentina provides university level in-service training*** The Chile count began in 1953; the Venezuela count in 1960.

i OTHERSTUDY AREA

...0 .. '" '".. ." III U ;>, .... II.... '" .. :l ;:;! .. 01 ....

m > '" OIl .... ... .. ~ 0 :l ....... ... III a 11 0 U N c::." .... 01 1lOITEM 41 III :l ... III

'" :l > .... .. Total .... 01 01... til ." III ... III 0 >< c:: 1lOIII c:: U .. c:: .... :l ... .... .... ...

:J! 41 ...:l .... a .... 0 '" 0 U 01 ,(I a III > <l:l <.:I ::c z u "" U <.:I "" IIC! ""

(1) Year service-began (19xx). 55 47 51 50 48 53 58 54 43 50 48 52 53 46 58(2) Did same .such service already exist? ('(es.No) N N N N Y Y Y Y ? Y Y N 6 Yes Y Y N(3) Year local officer became director 59 55 54 53 51 53 58 54 47 52 48 66 53 46 58(4) Year ervice was transferred to Host Government 59 .. 64 56 56 ., 58 60 61 61 .. 67 .. .. ..(5) Number of agencies at transfer 29 .. 15 9 32 .. 46 37 80 36 , . 22 .. .. "(6) Number of agencies in 1969 37 50 43 22 40 42 50 54 137 73 45 24 617 430 164 204(7) Rural population per agency. ( thousands) 90 40 47 51 28 18 196 70 51 41 67 61 62 35 17 31(8) Number of sub-agencies 0 0 0 0 2 1 0 0 362 0 25 0 0 0 0(9) Number of regional centers 0 0 6 0 8 6 8 20 12 7 4 0 67 + Chile 7 9 10

(10) Number of agents 37 50 43 22 42 42 85 54 142 73 460 24 614.+ Chile 700 192 191(11) Number of agents with Ingeniero degrees 0 0 0 2 25 4 85 48 142 6 220 0 312 + ChUe 430 5 180(12) Number of assistant ana rural youth agents 0 19 6 19 16 43 150 266 502 0 * 22 * 230 III(13) Number of home economists 8 30 0 11 18 39 50 25 42 31 0 20 274 270 230 140

(14) Number of supervisors for adult farm programs 4 1 6 2 8 7 9 5 13 7 IS 2 79 ? 37 16(15) Number of specialists in agriculture & 1ivestodk 3 2 6 3 17 0 0 18 29 14 * 5 79 + Chile * 30 35

+ Ecuador

(16) Training: orientation (weeks) 5 5 6 1 7 2 12 3 4 4 5 4 Average 5 ? 4 20(17) Training: in-service (weeks) 2 3 2 1 ? 3 ? 2 1 0 4 3 Average 2 4 4 *.. '(18) Doe8.ervice~attempt to cover all rural areas? N N N Y Y Y N Y Y N Y N, 6 No Y Y Y

(19) Number of local directors since row (3) year 1 3 5 2 3 4 9 7 6 5 5*"* 2 40 months 3 6**· 2average in-cumbancy

Source: Miscellaneous materials

TABLE 2

SALIENT CHARACTERISTICS OF EXTENSION SERVICES. 1969

48

A useful distinction is made between an organization and aninstitution. An organization can be an administrative unit perform­ing a particular job. It acquires the status of an institution, sup­posedly, when both the organization and the job become valued bysociety in general and/or groups within the society, and when thenormative values implicit in the job, and usually explicit in theorganization's program, become accepted and even professed byother groups. In short, the organization must have sustained andwidespread support. The support may come only from the peoplewho are served by the organization, and from a diffused constitu­ency attracted to it on moral grounds. Better yet, support shouldcome from other organizations "linked" to it either as suppliers,consumers, or joint sponsors of programs. In this way the organi­zation gets built into the society's institutional matrix, and is pro­tected by it. The organization itself must be sensibly structured,led by able people, and adequately staffed. It must be clear as toits job, and capable of programming an effective job -plan. If theseand other internal criteria are met, and the external linkages arestrong, then the IB model tells us that we are describing an institu­tionalized organization, and suggests some other evidence of in­stitutionalization that ought by then to be apparent, such as capacityto survive, influence, autonomy and the like. Some of the moreimportant criteria are applied to the extension program in thefollowing sections.

1. Access to Budget Support

Poverty is a problem common to all the extension services inthe study area, except the new service in Colombia. Typically, theextension service budget is determined by the ministry of financeand/or the president's planning office. These entities establishpriorities for public spending and allocate public funds among thecompeting ministries accordingly. The problem appears because(1) public revenue in the study area is notoriously low. taking asmaller percentage of GNP than is common in other parts of thedeveloping world; (2) finance and planning officers in the study areaattach low priority to all of the ministry of agriculture's programs(in Costa Rica, for example, the ministry of education receives 23percent of the central government budget and the ministry of agricul­ture receives 2 percent, a share for the ministry of agriculture notunusually small for the study area); (3) the principal constituency ofextension services, the small and medium scale farmers, is un­organized and has no political muscle (except in Chile and perhapsin revolutionary Bolivia and modern Peru); (4) the extension services,like mOE'lt other ministerial services, have been unable to securesupplementary domestic sources of finance; and (5) the extension

49

extension services were deserted by their original U. S. sponsorsin the early 1960s and have had difficulty in finding alternate sourcesof foreign aid since them.

Poverty means low wages, late wages, no fringe benefits, in­adequate gasoline allowances, jeeps with broken axles going with­out repairs, primitive visual aids and instruction pamphlets. Itmeans little fertilizer, seed, pesticides and other material to rundemonstrations. Most importantly it means no in-service trainingto speak of, and no agents to spare for training if adequate coursework could be organized. Government services in developingcountries are not often well financed, except those few that find awealthy and committed foreign sponsor. But extension is probablyin worse shape than any other. It is said that the conventionalextension operation is unusually expensive, in the sense that thematerial costs per professional, e. g., one jeep per agent, thoughtthough perhaps not necessary for efficient service, are high.When government revenue falls short of expectation, the ministryof finance cuts the material budget to the bone before it cuts thesalary budget, and the extension service has the most to suffer.

At the present time, several foreign financed, extension sub­projects, such as the BID and AID credit operations in Peru, thevertical wheat program in Bolivia, the cotton service in El Salva­dor, and the special AID supported "priority" zones in Panama,are relatively well financed. These exceptions are increasing innumber, undermining the argument about a "typical" condition ofpoverty. But from a continental perspective the latter is stillapplicable. It has worn the crowns off the once admired bite ofservicio extension, and reduced most agencies to a level of"pobreza" that allows for no initiative and no experimentation. 9

The salary situation is just as debilitating as the materialsproblem. Wages are low in the civil service, and good men areattracted away if and when opportunities arise in the private sec­tor or in those autonomous public agencies which are well endowedand can set their own salary scale. This attrition happens through­out government: not only to extension. But the extension agentposition is classified by civil service authorities everywhere at arelatively low rating, as it is generally assumed that a

9The cure for poverty may not be unlimited funds. One readercommented that abund<!-nce "may and often attracts the undesirable. "

50

personable, well motivated high school graduate can do the exten­sion job. Some writers share my views that the "extension job"is one of the most demanding in public service, and ought to drawa high rating. But the civil service authorities do not agree. Thusthe extension director can only offer modest salaries to beginners,though he wants to offer more. In some countries he can offerperiodic raises reflecting grade advances comparable to those inthe U. S. government. But these are infrequent - say once in fiveyears. There are no automatic within grade raises (except tocover inflation). There is a good chance at promotion in the firstfew years of service, because turnover rates are high enough toallow recruits assigned as assistant agents or youth club agents toadvance quickly to full agent status. But then they sit - - usuallyforever -- at a salary level that still compares unfavorably withmost other professional field agricultural assignments. There isnot much chance of further promotion, since extension is a bottomheavy service with lots of agents, a few officers and nobody inbetween. In Guatemala the veteran agents haven't had a raisesince the servicio expired in 1959.

The reader can imagine where he might go to find recruits fora fertilizer sales organization he was setting up. Not that theprivate sector is immeasurably wealthier, or always had positionsto offer the good agents. But generally it can at least double thegood agents' salary, and in Central America and several countriesof South America the private agricultural institutional sector isgrowing rapidly and picking the extension services to pieces.According to the IB model the ability to build effective links withresource-providing institutions are essential to the IB process.On these criteria the U. S. extension program in the study area wasan abject failure.

2. Image

Despite the budget problem, extension services have acquiredthe reputation of being rather popular organizations. Most of thepeople interviewed in Latin America who were not affiliated withthe extension service - - roughly two hundred persons, excludingfarmers - - felt that extension had not contributed much directlyto improving the productivity of agriculture. But many felt thatextension had contributed a great deal, indirectly, by introducingnew technologies to subsistence farmers and "opening the door"for latter-day change agents. And most felt that farmers and

•

51

farm wives appreciated the extension service and benefited from it.

Impressive evidence of the respectability of extension comesfrom responses to a questionaire. This was developed for thestudy, and passed out to about 150 American and Latin officials, offoreign and domestic agencies, who were not in extension but inpositions where having views on extension mattered. There aretwenty five questions in all (in English, which limited the size ofdistribution). All of them are direct or indirect tests of the imagecreated by the extension service about certain institutional factorstreated in the IB model. The questions were divided into fourgroups, entitled impact, leadership and prestige, relation withother organizations, and program. The evidence is subjective, ofcourse, which means that we are not learning what the impact ofextension has been but what the respondent thinks it has been (pre­sumably without the benefit of quantitative analysis). The ques­tionaire is shown in appendix b.

The responses indicate that the image of the extension serviceis considerably better than the assessment one arrives at from astudy based on direct observation of IB variable. Thus a pluralityof respondents feels that the work of the extension service "inimproving farm technology on small size farms which cooperatewith the agents" is good (outstanding, good, fair, poor and verypoor were allowed), and almost every respondent classified it atleast as fair. The latter comment applies for medium size farmersas well. The extent to which farmers appreciate the service, andtrust its technical advice, is thought to be good or fair. The pres­tige of the agents in the professional community is rated the sameway. These results are surprising, for although they are not incon­sistent with the often bitter denunciations of the extension directorsand agents, for pretending to playa role larger than life, that onehears from some respondents, they are considerably different intone.

Farmers themselves, at least those on the smaller farms inreasonably accessible areas, are disposed to identify extension asa primary source of acquiring new technology. I didn't interviewany farmers on this point. But Rosado and Laboy did, 1,162 ofthem, in a new and monumental field study of extension in 116

52

villages of Central America. 10 They asked about sixty extensionagents in the area to name two villages - - one receptive and onenot - - and they asked farmers in each village to name the naturalvillage leaders. Then they interviewed the leaders. They foundthat three -quarters of the leaders named the extension service asthe important source for introducing to them, and keeping them in­formed about, new practices. Although the results may have beenbiased somewhat by the interview design, the results do provideevidence, strong evidence, of a fa vOI'able attitude among the farm­ers about the usefulness of the extension services in CentralAmerica.

It is possible to say then that the extension service has estab­lished rather broad, diffused constituencies among a large clientpopulation and an interested professional public. Its image in thesequarters doesn't get the services any more money. But is doesmean that if they get more money, they would probably regain theremarkable renown and aura of success that they acquired in thegolden years of the servicios.

3. Functional Linkages

Aside from the diffused constituencies just mentioned, ex­tension services have established no strong links with otherinstitutions. This is really an incredible situation. It is onething to argue that the extension agent should not get personallyinvolved with credit or with the distribution of farm supplies.That argument can be defended. But it is clearly absurd forthe agent not to thoroughly familiarize himself with the activitiesof credit agencies and farm suppliers, and coordinate his pro­grams with theirs. It is abs,urd, that is, unless you insist thatthe extension mission in rural development is independent ofother inputs. Even then, if you believe in the Esman group'sIB model you know you have to make friends, fast.

The extension directors and the agents, however, appearto have avoided linkages. Human psychology provides some of

10Humberto Rosado and Maria Justina Laboy, Estudio deImpacto de los Servicios de Extension en el Istmo CentroAmericano (Publicacion Miscelanea No. 70; Guatemala: InstitutoInteramericano de Ciencias Agricolas, Zona Norte, June, 1970).

i..~

1

'~,(

53

the reasons why. The directors are part of an official com-munity where institutional pride, personal pride, and ambitionsfor empire frequently interfere with decisions. Americans andLatins alike claim that [this is truer of the study area than it is ofthe U. S., and that the element of orgullo, (pride) had to be acceptedas a fact of professional life. I met all the-extension directors,and have trouble labelling any of them sinister, spiteful or sel-fish. In fact, their ability to articulate the extension doctrineand the role of extension in rural welfare is impressive. Butit is understandable how men who grew up to importance in theservicio period, when extension appeared to matter as an inde­pendent force in development, were not prepared to re-examinethat position after the extension service was transferred to localauthorities, where the power of their American sponsors couldnot follow.

The agents may suffer from pride, as well. But a betterexplanation of an agent's inability to develop working relationswith professional colleagues is simply that many of the agentswere young, inexperienced, untrained persons, who naturallyfound it difficult and were therefore shy about calling for meet­ings with village elders, municipal authorities, fertilizer mer­chants and professionals of other organizations. Some of theseagents, and some of the others, were no better trained to bevillage-level workers, in the sense of coordinating institutionalinputs, than they were to be technicians. So, in the absence ofpressure from their supervisors, they visited the villages alone,and spend much of their time there talking to the only peoplethey could comfortably organize--the youth.

The most important missing link is between extension andresearch. Not that agents don't visit experimental stations, orthat subject matter specialists working for the research serviceare not invited to participate on farm walks and field days. Thishappens. But hardly ever, and certainly not enough to providethe farmers with accurage technical advice. If one studies sched­ules for annual extension conferences and for pre -service and :in-service training programs, one can appreciate better what anagricultural officer in one of AID's Central American missionsdescribed as the shame of the extension service: the dispropor­tionate time devoted to extension methods at the expense of tech­nology. Since therels very little~ganyway, one can guesshow small must be the_absolutrlevel of training the agents get

54

in substantive matter. A way to compensate is to recruit a largestaff of subject matter specialists into the extension service, pro­fessionals with expertise who are free to travel. But this has nothappened yet to any significant extent in the study area, as thereader can verify by referring to table 2. Rosado's study of exten­sion in Central America shows that the practical effects of a smallsize subject matter staff is to provide the average agent with fewerthan five visits per year. This hardly compensates the agent's owninadequacies.

The question arises as to whether the Americans were re­sponsible for the isolation of extension services from other in­stitutions, particularly ones functionally related. The answerappears to be yes and no. In the servicio period, the extensionservice was linked to the research, credit and promotional d i­visions within the servicio to the extent the divisional chiefs-­who were American at first- - coordinated their activities andforced their staff to do so as well. Many of the Americans inthe research divisions had been transferred to the servicio pro­grams from OFAR, and a few brought with them a barely con­cealed contempt for the extension types. This, and the apparentinclination of some entomologists, plant pathologists, agrono­mists and the like to lose themselves in their experiments onthe station, partly accounts for the complaints one hears frommany of their extensionist colleagues within the servicio thatthe latter weren't getting sufficient support from the subjectmatter specialists. On the other hand, some of the U. S. ex­tension advisers preferred to be left alone to get on with theextension job, and didn't notice or mind the absence of thespecialists. These attitudes were not characteristic, however.In some servicios the American director insisted on close co­ordination between his divisions. Norman Ward had a reputationfor this in Ecuador, and Bailey Pace did as much in Honduras.In other servicios, the American extension adviser himself en­couraged close working relationships between his agents andthe subject matter specialists. Nick C. de Baca's efforts inEI Salvador and Guatemala and Ernie Gutierrez' work in Boliviaand Peru are good examples.

The servicio, in fact, was a good device for ensuring co­ordination between glamorous, well supplied units that mightotherwise fly off in different directions. Frequent staff meeting,joint programming, and the leadership of individuals who weregenerally concerned about efficiency, helped orient extension,research, credit, seed production, machinery, livestock andthe other staffs in the same direction.

? .

55

The problem came later. on the transfer. Extension serviceswere shifted out of the cooperative. servicio environment. usuallybefore the other divisions were transferred, into a ministerialstructure which included some functionally competitive divisions.some hostile and spiteful bureaucrats. and very few officers whounderstood or sympathized with the extensiondoctrine. The linkages which had developed within the serviciowere broken. and nothing had been prepared to replace them. Thetransfer process itself was at fault. This. of course. calls intoquestion the entire servicio concept since it would appear to havetaken a prodigious job of institutional engineering for a servicioextension chief to develop links within the servicio. and at the sametime prepare links within the ministry. when the servicio and theministry were at odds. One solution would be to let the servicioabsorb the ministry. as happened in effect in Peru. Another wouldbe to transfer everything together to a new domestic institution out­side the ministry proper. This happened in Honduras. In bothcountries. links between extension and other servicio units werepreserved (the extension- -research link in Peru had never beenstrong. even in SCIPA. the servicio). But such happy solutions ofthe transfer problem were generally not offered by the host govern­ment.

4. Doctrine and Program

There is no doubt that the institution builders had a clearlydefined doctrine. and a set of programs consistent with it (wewill discuss the merits of the doctrine later.) There was an agree­ment within each extension service--between directors, super­visors and agents--on what the extension role was and how theextension process should be performed. There was agreementwithin the international community of extensionists as well,owing to: (l) the commonality of experiences and training ofU. S. advisers; (2) the enormous attraction and influence of asingle U. S. academic institution, Cornell University, as the pri­mary training ground for graduate Latin extensionists; (3) theefforts of the U. S. Federal Extension Service to guide the develop­ment of extension services throughout Latin America; and (4) theemergence (in the 1950s) of the powerful Inter-American Institutefor Agricultural Sciences (IlCA) extension team at Turrialla,Costa Rica- -particularly as it was shaped by Joseph Di Franco,who was on that staff from 1958 to 1968--which codified extensiontechniques for all the Americas and pressed them vigorously in

56

publications, in classrooms, and on tours around the continent.Deviants were ostracized, and deviant cOlIDtries were literallyomitted from the map. This happened to Tito Valle, former headof the Honduran service, who tried to argue at a 1961 meeting ofextensionists in Mexico for a new focus for extension and wasshouted down. The Honduran service is still not accepted by theold guard as extension, and you can find maps of extension in LatinAmerica which omit the name of that country. The new rural devel­opment service in Chile is treated the same way ("Chile doesn Ithave an extension service any more"), though the "agriculturaltraining" chiefs I talked with in Santiago in December 1969, whileadmittedly successful in submerging the extension function withina broader, development, agency (SAG), seemed to be able mentrying to make the extension idea work, and they had retained mostof the agents and agencies inherited in 1967. The point here is thatthe U. S. extensionists who originally shaped the extension organiza­tions in the study area probably had as clear an idea of where theywere going as any institution builders)1ave ever had, and impartedthis wisdom and confidence to the staffs they left behind. So wemust give a high IB rating on these criteria.

There are some qualifications. First the program is at faultfor not sufficiently emphasizing evaluation and feedback. These arethought to be vital processes in any institution, and extension ser­vices never have done enough of them, or at least never have donethem the right way. Agents were required to submit reports. Head­quarter staffs put the reports together. Extension evaluation teams -­many of them sponsored by IlCA--assessed the reports, interviewedagents and farmers, and made recommendations. But the issuesthat were analyzed were mostly organizational, the results thatwere measured were (and still are) mostly the number of farmercontacts. Rarely, if at all, did anyone look to see whether agentswere having an impact on production, whether youth leaders wereleading youth, whether house wives were influencing the familybudget.

Second, the program is at fault for not sufficiently emphasizingtraining. This has been a continuing problem for extension sincephase-out. The IlCA evaluation studies hammer at this weaknessrelentlessly, much to their credit. There is practically no pre­service or in-service training provided by the extension services.What little there is concentrates on extension methods and communi­cation theory. Even assuming the correctness of the extension doc­trine, the poor preparation of the field staff for working with far­mers was so obvious that, in retrospect, one can insist that the

57

Americans should have ensured that training was part of normaloperations and safe from budget cuts. The problem was exacer­bated by the extremely low level of human resources available forhiring, especially degree level professionals. The U. S. Coopera­tive Extension Service since the mid 1930s has recruited only de­gree holders. The Latin American services, except for Chile,Peru and a few other countries, can't find them, or buy the fewthat are on the market.

5. Other IB Variables

We have looked at some of the important parts of the IB model,including what are called the institution variables (internal structure,resources, doctrine and program) and the linkages. One otherimportant institutional variable will now be introduced. This willbe followed by short discussions of some of what Esman calls thetests of institutionality.

Leadership. IB refers to this as the most critical element ininstitution building "because deliberately induced change processesrequire intensive, skillful and highly committed management bothof internal and of environmental relationships." It was difficult toget a good hold on this variable, since my visits with the leaderswere too short to form impressions about their political viability,professional status, technical competence and organizational com­petence, all of which are elements of leadership emphasized in theIB model. In general, one is favorably impressed by the directors-­by their dedication to the extension doctrine and their ability toarticulate their problems. They were generous with their time,and helpful with my analysis.

It was nevertheless possible to assess one of the leadershipvariables, --continuity--and the results are impressive. Theability of extension directors to hang on to their jobs, despite majorpersonnel changes at ministerial and even higher levels, is remark­able. The servicios were famous for supporting continuity ofleadership in an otherwise fluid environment, and the practice hasapparently been tolerated since transfer. It is a mixed blessing,since the leaders and their doctrine can ossify.

58

Survival. The services appear to have survived the firstdecade since transfer. In no country has the national extensionservice been terminated altogether, although in Colombia it wasshifted out of the ministry where the Americans left it and reorga­nized. But this achievement may be illusory. In a few countriesthe organization is almost completely isolated from government'sdevelopment program and the field staff appears to be just goingthrough the motion of carrying out the extension job (though notthe headquarters staff, which everywhere is busy). The serviceas a whole may thus be no more than a sinecure for a beaten leader­ship and ill-equipped staff, except, that is, for those agents whouse the experience as a base for better jobs elsewhere. Esman hassaid that survival in this sense is the mere shadow of the importanttest of institutionality.

The conventional extension approach is being questionedeverywhere, and one can predict major changes in leadership andprograms in the next few years. This is already true of Colombiaand Chile. In Honduras, the post-servicio agricultural authority,Desarrural, has jost lost its autonomous status. In Peru andParaguary the services are being forced by the administration totake on agrarian reform and other unfamiliar jobs, and the res is ­tance of the old guard appears to be futile. In Guatemala, thereare some doubts whether the extension service will be invited toparticipate in the new, agricultural sector program, and there istalk of creating a new field staff of action agents. In Costa Rica,the service is being forced to regionalize and tighten relationshipswith other organizations, in order that the government may secureforeign credits that would otherwise be withheld. In Bolivia, someof the agents are being drawn into the vertical crop and livestockprograms, where they are better equipped to operate, etc. Exten­sion is in a state of flux, and new forms are replacing the old. Infact, in Chile the word extension has been deliberately omittedfrom the title of the reorganized extension service, because theplanners want to break the spell of the U. S. model over localactivities.

Autonomy. Extension services have no autonomy, in theimportant sense of having control over budget levels, budget alloca­tions, salaries and other financial affairs. Honduras and Peruused to be somewhat anomalous, because in both countries theservicio structure survived phase-out and the successor institutionretained considerable authority over the disposition of funds. Butboth organizations have now been broken, and the pieces integrated

59

into the ministry.

Extension services have had lots of autonomy, however, inprogramming their activities, a practice which apparently has donemore harm than good. Program autonomy is probably undesirableand not a criterion of successful institution-building in this case.The issue arises also in the struggle of extension directors forequal ranking with research and agricultural education units in theministry. In some countries, all three units, and a few more, areequidistant in the organization chart from the central directorate ofagriculture. In Ecuador, extension was in effect put above the fieldservices of the other units. The agents became the delegados ofthe ministry, and livestock and other specialists report to them.In most countries, however, the extension director ranks belowthe research director, or the fomento director. This apparentlycreates an intolerable psychological strain. The ex-director ofthe Paraguay extension service fought for two years after transferto the ministry in 1967 to gain equal bureaucratic footing with re­search. He got it finally, but then lost it while he was travellingto Europe on a short training program.

Extension's claim to equal rank is based on the conventionalmodel, and the validity of the claim depends on the validity of themodel. At first I was prepared to give high marks to the extensionservice that had a superior position in the ministry structure.This is what the IB model would indicate. I now think that is amistake.

Influence and Spread Effects. Extension services probablyhave scored their greatest success in this area. The concept ofproviding advisory services to the farmers was generally unfamil­iar to public and private authorities in the study area before BAAprograms were initiated. Now the extension function is widelyacce pted as an important phase of any rural development program,and the work of the regular extension services have been imitatedand improved upon by a host of other organizations. The regularservice is the only entity which limits itself to extension. Else­where extension staffs are built up within sales and service organi­zations that are commodity or input oriented.

Beyond that, there is some evidence that the more funda­mental notion of a government's responsibility to service its ruralconstituency has made inroads on traditional bureaucratic thinkingin Latin America. This would be extremely important effect ofAID's institution building efforts. It is helped by the fact that somany of the professional agriculturists in Latin America -- including

60

many of those at the pinnacles of leadership--were trained asextensionists during the servicio and post-transfer periods. Onemight even be able to defend the entire U. S. extension program onthe grounds that it has created a training ground for professionalswho then found permanent and productive employment in other units.

6. Trade-offs

The IB model does not tell us how to evaluate an organizationwhich is weak in certain respects and strong in others. It empha­sizes a few of the institutional qualities, but doesn't predict howmuch of one can be sacrificed without loss of institutionality.

There is some consensus in recent literature that linkagesare the dominant variable. This was the conclusion, for example,of the rapporteur at an AID conference on institution building heldat Purdue University in 1969. 11 It would seem axiomatic that link­ages are absolutely indispensible to extension institutions: thelinkage concept is built into the work extension, and the idea ofcarrying on an extension job without backward and forward linksappears definitionally inconceivable. If this is so, then there is nopossibility of tradeoff with respect to linkages, and the fact that theextension services in the study area fail the linkage tests in mostimportant respects would appear to be sufficient grounds for failingthe entire institution building process, the commendable achieve­ments in other institutional tests notwithstanding.

Tradeoffs may be possible, however, in assessing the over­all significance of the U. S. extension program. Even though theextension organizations that grew up within the servicios failed toachieve institutionality, nevertheless the other effects of theunsuccessful institution building program, in particular the influ­ence the program had on farmer services offered by other organi­zations, could compensate for that failure. I do not have enoughinformation to make such a judgment. Theoretically, extensionservices could fail institutionality tests and still have substantialimpact on farmer productivity. If that were so, the choice

11George H. Axinn, "Summary Report" (Summer Workshopon Agricultural College and University Development, July 27­August 8, 1969; Lafayette, Indiana: Purdue University, 1969Lp. 18.

61

of labelling the U. S. effort a "failure" or a "success" would berather difficult. But the problem is academic, because that hasdemonstrably not been so.

D. Adoption versus Adaption

A common complaint about all technical assistance projectsis the failure of the foreign advisors to adjust a foreign technology orinstitution to local circumstances. The popular way to put this is tosay that the advisor knew how to adopt but not how to adapt. Onehears this point repeated again and again in the twelve countries byboth professed friends and professed enemies of extension and theU. S. In Spanish, the words are also close--adopcion, adapcion.

The majority of the criticism about the American's ability toadapt to Latin American circumstances was silly, trite or unfair.There was a story about the U. S. home economist in Ecuador, whotaught her ladies how to preserve apples, despite the fact thatapples grew the year-around. There was a story about an extensionagronomist who insisted on growing sorghum for people who hatedits taste. Usually the problem arose because of such a differencein taste, or because the technology was not appropriate. It is saidthat Americans didn't study the region before starting projects, andif they had, the problems could have been avoided. But I am satis­fied, after three years of looking at the subject, that the issue thusstated is exaggerated and that most Americans made an effort to in­form themselves. They rarely blundered about with blinders on.These critics are probably bothered by something else, and areusing the adapcion argument because it is convenient.

A few critics cannot be so easily dismissed. They are lessconcerned with the U. S. technician's ability to adjust taste andtechnique to local life. than with the fundamental concepts whichguided his action. These critics ask whether the extension modelis appropriate. whether the land grant idea fits. whether 4 -H clubsmake sense. whether all farmers need to approached and whetherthey need be approached as individuals rather than in groups. Onthe larger issues the American advisers were inflexible. whichhelps account for the extraordinary similarity in structures andprograms of extension services that persisted until recently. Thus.although the usual argument about adaption can be deflected, theproblem -can be formulated in a way of transcending importance.

62

E. Education versus Fomento

There is a plaque overhanging the desk of an officer of theConsejo de Bienestar Rural (CBR). the Venezuelan semi-officialextension training agency. It quotes one of the elder Chineseextensionists. Kuan-Tzu. who was concerned back in the ThirdCentury A.D. about the human resource problem:

If you plan for one year. plant wheat.If you plan for a decade. plant trees 12If you plan for a lifetime. train and educate the farmers.

The decisive weakness in the U. S. extension model is theemphasis it gives to its own peculiar interpretation of education.The extension process is an educational process. it is said. Thisprincipal is enunicated early in every major textbook on extensionprinciples. It is based on the unassailable assertion that to getfarmers to change their behavior- -to adopt a new practice. ratherthan simply apply it. temporarily. while favorable conditions arti­fically contrived by a change agent prevail- -the farmers must learnhow to do it by themselves. and the reasons for doing it. Thefarmers must be taught to "solve their own problems." to "helpthemselves. " to "make decisions. " to "choose. "

The model forbids the agent from doing the job for thefarmers. That is reasonable. But the model also outlaws allelements of coercion. and the substitution by the agent of unfeltneeds for felt needs in the planning provess ("how to think. not whatto think"). The model. as it has been interpreted in the LatinAmerican scenario. would also appear to outlaw the very act ofcreating favorable conditions that might induce earlier adoption-­making credit and supplies available. developing a market. artifi­cially structuring prices. etc .• - -and one supposes these are notacceptable because they appear to intervene in the educationalprogram.

The focus of an educational program ought to be on the learn­ing process. If farmers do learn new practices under programsinvolving some coercion. if they do learn to use and appreciate

12Translation from Spanish by the author.

f

t~I·..'.:

;1

l

~

63

technologies they were unable to identify as a felt need, if they doadopt unfamiliar systems within a specially protected economicenvironment, then the educational program has accomplished itsobjectives. And they do, time and time again, as the studies ofproductivity show. The diffusion of innovations occurs more rapidlyunder these contrived, non-permissive, regimes than it does underthe pure extension method. And behavioral changes are demon­strably permanent. If they aren't, it usually means the economicenvironment was unfavorable and the farmers were well advised torevert to previous practices. Thus the farmers are learning underthese unnatural conditions; though they haven't been trained withapproved methods.

The professional extension school seems to have gotten hookedon a method, and lost sight of the objective. The school has wronglypretended that the choice ~s between education and promotion--whatthe Latins call fomento. 1 That is a wicked formulation of the issue,because fomento includes an educative element, and it may evencall, on occasion, for the "extension method." True, it does tendto use less of the permissive instruments in the extensionist's bagthan the manipulative devices that are associated with an "elitesttheory with an implicit social engineering bias." But if properlydone, fomento will educate, nonetheless. The extension school hasno monopoly of the learning process, and its own methods, unlesscarried out within a fomento program, are terribly unproductive inthe typical subsistence cultures of Latin America. The next chaptershows this. That is why some of the new people in Latin Americanextension want to rename and redefine the extension function, andlook for quicker ways to get information on farm technology to thefarmers in a way that they use it. That is why some of the newpeople in Latin American extension want to upend the standard doc­trine, and instead of advocating "educacion por accion," deviseways of "accion por educacion. "

A related issue concerns the proper selection of extensionprojects. The extension model prefers that villagers determinewhat they want to learn. Under a planning system known as "pro­gram building," the agents sits with the elders, encourages themto identify their needs and helps them determine a course of action.

13The word means what it means in English: to promote, tofoster, to encourage, to foment.

64

The difficulty with program building is that subsistencefarmers can neither identify nor solve their real problems.Not even their leaders know the alternatives, and they won'tlearn them from the friendly chatter of an uninformed changeagent. They have to get involved in the new program, or observea neighbor who is involved. While it is true that result demon­strations are featured in the extens ion approach, the sort ofmass ive demonstrations that are needed to get widespread adop­tion are in themselves a fomento activity. They are not designedby the farmers.

It is inconceivable that village elders in Asia or LatinAmerica would originate a request for a short, stiff-stalkedhigh response wheat. What they are mor e likely to ask for, asa former director of the Ecuadorian extension service once said,with only a trace of a smile, are more park benches and shoes.These are the sort of problems which they can hope to solve.The other solutions are not conceivable. The farmers and theirleaders are no more likely to ask for a dwarf wheat, as a solu­tion to their problem with lodging, as for the wind to stop blowing.

There is no denying that the small farmers of Latin Americaare rational, wise people. Farmers everywhere are usuallywise. But they don It know which constraints can be relaxed bypurposeful human intervention and which cannot. Moreover,the economic environment is literally "out of (their) control. "If all the elders in all the villages selected increased corn pro­duction as top priority, and the extension service played theirgame, the increase in off-farm deliveries could not be absorbedby the market. None of us would like to be associated with theresulting disaster. The improved practices we want to encouragein Latin America generally involve purchased inputs and cashcrops. Their viability depends entirely on the market. Thevillage elders cannot manipulate that. A fomento agency can,however. Program building should be interpreted as a way toencourage farmers to accept a development program alreadydescribed on a sheet stuffed in the change agent's hip pocket,not as a way to develop the program. This is what the Honduranswere arguing ten years ago, when the old guard threw them outof legitimate society.

65

What appears to have happened in the extension movementin Latin America is what the institution building profession callsII goal displacement. II

"Men and organizations are simply not capable ofdeliberately acting indeterminately. Some sense of aimor purpose, some sense of meaning, must underlie humanaction. In complicated, organized action that runs overtime, the need for a sense of goals and purposes is asimportant as it is obvious.

Given this need, what happens when a problem cannotbe defined in technological terms, and where the apparentrelation between intervention and desired end-state con-d ition is indirect and even somewhat obscure? A lotof things may happen. They range from "lucking out "_­achieving a goal without ever understanding how- - toutter and abject failure. Often, however, what happensis "goa1 displacement ". Individuals and organizationstend to fix upon rather concrete, specific, measurableforms of activity--and to equate such action with thegoals for which it is assumed to be instrumental. Andthis is dangerous. "14

It is particularly dangerous when the ability of the institu­tion to achieve the original goals is itself doubtful, and the staffnever realizes what it is missing. Self analysis becomes dis­torted by a persistent tendency to define extension in organiza­tional rather than functional terms.

F. Comparisons betwe.en Countrieswithin the Study Area

Originally, I had intended to compare the twelve extensionservices with one another, using various indicators of institu­tional viability. I hoped to rank the twelve by levels of achieve­ment and then see if there was any significant difference betweenthe technical assistance programs that were associated with theextension services at the top of this list, compared with the

14W. J. Siffin, "Institution- Building for Development andSocial Change: An Evaluation of the Perspective" (Lafayette,Indiana: Purdue University, August 1969), p. 11 (mimeograph).

66

technical assistance programs that were associated with thoseat the bottom. This might have provided some clues about thepreferred way to assist extension organizations in particular,and institutions in general.

The project was largely abandoned, for two reasons.First, there was no apparent dispersion of the twelve serviceson the success scale wide enough to easily distinguish good frombad. This may be due to the study's faulty sensing devices. Itprobably has more to do with the ubiquitous budget problem andthe prevailing doctrine. Both of these have left most of thesix hundred agents in the study area without meaningful employ­ment. Second, it began to appear that the only successful ex­periences would be those that had drifted off the mark, and abusedthe transplanted model. If this were so, one would have to givehigher grades to those technical assistance programs which wereleast successful in accomplishing the purposeful transplant ofthe model. This was possible, but it would have led to an oddset of questions.

Despite these problems, some evidence was available tosupport rough inter-country comparisons, and a modest stab atthe project was made. The methodology and findings are presentedin the general report.

G. Comparisons with other Countries

The collective image of the extension services of Mexico,Venezuela, and Argentina is far superior to the collective imageof the extension services of the study area. Brazil has a betterreputation, too, though it is possible that that is based on only afew of the state organizations and other state organizations wouldnot compare favorably.

The extension division of INTA in Argentina has an out­standing reputation. Ignacio Ansorena, an extension specialistof Uruguayan nationality based at Turrialba (the only survivorof the Di Franco team), used to cover the IleA Southern Zoneand is familiar with the Argentine experience. He says the repu­tation is deserved. He attributes it to many factors. Chief amongthese are (1) the inseparable links with research personnel, whoare in charge of the whole research-extension process; (2) theemphasis on training, which includes periodic year-long sessionsfor every agent at the new Graduate School in Agricultural

..~

;

.,'1

67

Sciences at Castelar (and which offers, incidentally, a mastersdegree in extension); and (3) the prestige of the director of ex­tension (Reichart) and his deputy (Castelli) and the fact that theyhave been associated with the program since INTA was formedin 1958. We would have to add the fact that INTA is financed bya 1 1/2 percent tax on agricultural export earnings, which, inmost years, enables it to fully equip its agents and to pay attractivesalaries. Also, it is dealing with more sophisticated and wealthierfarmers than extension's clientele in the study area. In somedistricts, the farmers have begun to contribute to the extensionbudget.

The Argentine phenomenon is not all favorable. Budgetcuts in 1969 attributable to a fall in exports forced the INTA ad­ministration to freeze salaries and employment, and led to theloss of many good technicians. Moreover, the extension directorsadmit to being heavily influenced by the U. S. model, and take ex­ception to the proposal to give extension a fomento role. Theyprefer to keep extension pure, which conflicts with the conclu­sions of this study. Nevertheless, it is important to recognizethe prestige of the INTA service and the fact that it has suc­cessfully altered some of the less desirable aspects of the con­ventional, transplanted model.

The extension services of Brazil were not included in thisstudy, and cannot be properly assessed here. ACAR of MinasGerais is highly rated by its observers, of whom there are many.A few other state services have acquired international reputations,as has the federal organization ABCAR. Americans under con­tract with the Rockefeller technical assistance organization, theAmerican International Association for Economic and SocialDevelopment (AIA), helped start ACAR in 1948, and ABCAR in1956. FOA personnel got involved in 1953. Thereafter the offi­cial U. S. contribution grew rapidly. By 1962 there were sixteenU. S. extension specialists identified as such on the AID staffingpatterns for the Brazil Mission. This was the high water mark.In 1964, the number of agencies in Brazil surpassed for the firsttime the number of agencies in the study area (425 versus 408,excluding Chile and Panama), and has grown at a much greaterrate since then (:1,078 agencies in Brazil in 1969, for a rural popu­lation of 49, ODD, 000 versus 530 agencies in the study area, ex­cluding Chile and Panama, in the same year, for a rural popula­tion of 34, ODD, 000). One could design an impressive comparativeevaluation of extension organizations simply by studying Brazil'sstate services.

68

One characteristic that distinguishes the Brazilian ex­perience from the study area is that the financial base is sharedby the state and federal governments. A second distinguishingfeature, and one that might explain the greater vigor and reputa­tion of extension in Brazil, is the fact that it has always been tiedup with supervised credit. Roughly half the farmers contactedby ACAR agents receive credit from them. Another character­istic is that the Brazilian services concentrate their work. Theyclaim contact with farmers in only 26 percent of the municipiosin the country (1,043 out of 3, 965).

The principal difference between extension services inMexico and Venezuela, on the one hand, and those in the studyarea, on the other, is the comparative wealth of the former group.With a large budget any service can gain respectability, just asthey did in servicio days. The recent financial history of theMexican extension service is a case in point. The 1969 budgetwas five times as large as the 1966 budget, and it was supposedto almost double again in 1970. The increase was due first tothe large contribution that state governments began to make.Second, and more importantly, it was due to the contributionsthat began to flow in from the Office of the President, from acontingency account at the disposal of the President for specialprograms. The number of agents has tripled since 1966, andthe expans ionist plans boggle the mind. Jeeps are in good supply.And funds arrive on time. The key element in the Mexican ex­perience is the personal interest President Diaz took in the ruraldevelopment program. The leadership in the Office of the Presi­dent obviously concluded that the extension service should getthe job of spear-heading the assault on rural poverty. Resourcesand commitments of this sort would be rated highly by the IBmodel. But one becomes uneasy about the Mexican example.The funds are pouring into a service which is still orientedon traditional lines. Linkages in all directions appear to be weak.A good bet is that if the linkages are not strengthened, and if thedoctrine is not adjusted to incorporate more of the fomento ele­ments previewed in the Puebla Project, a pilot research-cum­extension program for subsistence corn farmers run by CIMMYTin the state of Puebla, the Mexican extension service will dis­appoint the President's officers.

If we consider that the budget factor is a variable thatdoes not fairly measure the success of the institution buildingeffort,and ignore it, then the significance of the Mexican andVenezuelan experiences in the comparative analysis is different

1

j'~

69

than it otherwise appears. It indicates that extension servicesin two of the control countries developed in much the same wayas in the study area. Thus the official American teams did notpervert the development of extension services in Latin America,by imposing a doctrine and methodology that was unpopular. Ex­tension would have developed gradually along these lines anyway,since the American model was universally accepted as correct.Venezuela paid for its own American consultants. And Mexicojust looked across the border.

Argentina broke this pattern. It studied the hemisphericexperience during the mid-1950s and planned a different type oforganization. Again, disregarding the budget factor, it wouldappear that the important lesson of the control countries is theone provided by Argentina and Brazil. It is entirely consistentwith findings elsewhere in the study. It is, simply, that exten­sion is a tool, and tools ought to be used by institutions but notinstitutionalized.

i~ .

'~;

I

'I

CHAPTER 4

SIGNIFICANCE: INCREASING AGRICULTURALPRODUCTIVITY

A. Productivity and Other Significance Effects

The author does not argue that the principal, declared objec­tive of extension services has been to increase the level of farmproductivity, and, thereby, the level of rural welfare. Economicobjectives are invariably included among the extension services'priorities, but occasionally they are ranked below such "non­economic" targets as community development, home improvement,leadership training, etc., all of which may be considered othersteps to the welfare goal. A review of the literature of the late1940s and early 1950s shows a clear stress on economic factors.But a sampling of opinions among the Americans who designed theextension programs in Latin America would show more mixedresults. Some would have emphasized production targets. Otherswould have emphasized the social advantages. Most Americansprobably assumed that the development of extension organizationswould serve both purposes. The author believes that increasedproductivity and production of marketable goods ought to have beenthe principal objective, since remunerative employment seems tobe the most powerful engine of a development process that, prop­erly managed, offers to achieve all these objectives. But hedoesn't defend the point. The readers who are otherwise per­suaded are asked simply to accept that productivity is an impor­tant objective, and that an evaluation of the relation between exten­sion and productivity is worthwhile.

This chapter deals exclusively with the extension-productivityrelationship. Moreover, it deals only with the impact of LatinAmerican extension institutions on productivity, and ignores thedirect effect on production aU. S. technician may have had throughhis own extension activity. The latter is not an effect of the in­stitutions the U. S. program created; it is a significance effect ofthe technical assistance program nonetheless. With this omis-sion I do disservice to many excellent technicians, whose storiesought to be told.~'<

*See next page.

71

72

There are other significance effects which will also be ig­nored in this chapter. For example, there are the indirect effectsof extension institutions on productivity, and an excellent exampleof these would be the contributions attributable to their extensiontraining made by ex-extension personnel who have moved on toother jobs. There are the effects on significance indicators otherthan productivity, and in this perspective the leadership training,home improvement, decision making capabilities and other impor­tant non-agricultural objectives of conventional extension educa­tion systems would deserve attention they are not given here.

>:< A list of the most effective Americans involved individuallyin direct extension type activity in the study area would include(but in no way be limited to), if we can accept the word of theircontemporaries in AID and/ or the host government officials theyworked with: Johnny Johnston and Fran LeBeau, for their workin agronomy in Guatemala; Ben Birdsall, for organizing the mas­sive demonstration program in EI Salvador; Bill Conkle, forhelping the extension service ll get involved" in Nicaragua; HaroldMowry, who helped develop the coffee research and extensionservices in Costa Rica; Jim Bleidner, for his work with livestockfarmers in Colombia; ll Tiny" Ward, Charlie Padgett, RussDesrosiers, Bob Smith, and "Spud" Bullard for work in Ecuadordescribed in the general study; R. D. Conover, for successfullyintroducing dairy farming practices in the Arequippa area, Peru;George York, who helped eradicate the umbrella ant in Paraguay.The best of them all, with a reputation that extended far beyondPeru, was the remarkable John R. Neale, who died just beforethis study began.

Among the technicians I met, my favorite is Ray Stadelman.He has roamed these hills since the late 1930s (making him thesenior officer in the select corps), working on tree crops, rootcrops, grain crops and bugs, sociological studies of village life,extension handbooks, flood repair, snakes and the rest. He hasmid -wifed at birth of babies in villages north of Huehue, extractedabcessed teeth, bombed a motorcycle straight up a volcano, anddriven me out of mud two feet deep. He can repair anything in hishome workshop, which he carries everywhere despite the factthat it uses up most of his freight allowance. He ought to be everyman's all-American "ugly arne rican11 • When he reaches 70 years,in a few years, I think he intends to take his marvelous Mexicanbride, his motorcycles and his shop, row across the Pacific, andstart again on an unfamiliar shore.

73

B. Design of the Productivity Study

For the purposes of this paper, the plan is simply tosketch the outline of the field research program, and then passdirectly to the overall findings. The full report includes severalhundred pages wherein the methodology, description and resultsof each field study are presented. What may appear to be norma­tive judgments in the following section are supported by the factspresented in that report.

The field research program was divided into three phases.The first two comprise what I have called the "impact" study.Nine areas, ranging generally between 200 and 2,000 square miles,were selected for an intensive review of the regular extensionservice's contribution to whatever improvements in farm tech­nology have occurred there. The principal criteria for selectingthese "study districts" were (1) that they have been settled, forgenerations, by small scale farmers engaged primarily in theproduction of basic food crops; (2) that some observable improve­ment in farm technology has occurred in the last twenty years, atleast for one crop and at least in one part of the district; (3) thatthe extension service has been busy in the district for a longperiod, and by virtue of the length and continuity of that work, ora special emphasis given to the district, appears to have contribu­ted to the observed progress. In short, I was looking at nine smallfarmer areas where extension has been active and progress hasoccurred, and trying to see whether extension had indeed con­tributed and how much. There follows a short description of eachof the nine districts.

Quezaltenango, Guatemala. The Western Guatemalan highlandssurrounding the city of Quezaltenango, including an area ofapproximately 350 square miles and a rural population of 160, 000.The area lies between 6, 000 and 9, 000 feet and is predominantlysettled by Indian small farmers who grow mostly corn and wheat.The district includes hilly land to the north and east of the mainvalley, but no village is more than three hours from the city bytruck. There is one extension agency, located in the city.

Jutiapa-Jalapa, Guatemala. The triangle formed by the threetowns Jutiapa, Jalapa and Asuncion Mita in eastern Guatemala, in­cluding the areas covered by the extension agencies located inthose three places. The area is of approximately 1,100 square

74

miles with a rural population of 140,000. It lies between 1,500and 4, 500 feet, except for the Montana mountainous area west ofJalapa where villages are as high as 6,000 feet. Corn, beansand rice are the basic food crops, though some wheat is grown bythe Indians on the Montana. The farmers are mostly mestizosmall holders. There are four extension agencies, including theoccasionally vacant one in the town of Monjas.

Northwestern EI Salvador. A large, densely populated areastretching from Ilobasco on the east to Ahuachapan on the west,and from the frontier of the coffee zone on the south to the LempaRiver on the north. The area comprises 1,250 square miles and550,000 rural people, and lies between 1,000 and 2,000 feet. Itis mostly cultivated with basic food crops, particularly corn,beans and rice. There are some large farmers, but the majorityof the land is given to small scale farming by mestizo farmers.There are twenty-one extension agencies in this large studydistrict.

Boyaca, Colombia. The highlands of the department of Boyaca,ranging in altitude from 7, 000 to 11, 000 feet with an area of2,250 square miles and 450,000 rural people. These highlandshave been extensively farmed by small holder Indians since beforethe Spanish occupation. It is a potato culture, with importantwheat and barley crops, of post Colombian origin, as well. Novillage is more than a few hours ride from the important marketsat Sogamosa, Tunja and Chinquinquira. There are now twoagencies in the study area, but during the 1950s the Americansestablished eight agencies there in a pilot extension program.

Azuay-Canar, Ecuador. Two economically depressed provincesin the southern sierras of Ecuador, centered around the cityCuenca, and relatively isolated from the mainstream of develop­ment in the country. The district includes the settled parts of thehighlands, an area of approximately 2,000 square miles and arural population of 450,000. The area lies between 7,000 and9,000 feet, is predominantly settled by Indian and mestizo smallfarmers who grow potatoes and hardy cereal crops. There is atradition of supplementary, handicraft employment in manyvillages. The extension service has five agencies in the area, anda regional development authority has another thirteen.

.J

..~

•

75

Mantaro Valley, Peru. Those parts of three provinces of thedepartment of Junin which include the densely settled highlandbasin and surrounding hills. The valley is the highland settle­ment closest to Lima, and is occasionally referred to as thegranary of the capital. The chief city in the valley is Huancayo.The basin is 10,000 feet high, the district covers almost 500square miles and a rural population of 300,000. It is farmed bysmall holder Indians and mestizos and a few large farmers.Potato is the principal food crop. The extension service has azonal headquarters in Huancayo and five agencies and about thirtysub-agencies in other parts of the valley.

Puno, Peru. The northwest shore of Lake Titicaca, with 6,000square miles, 550,000 persons and an altitude between 11, 500 and12,500 feet. This is a small holder, Indian potato culture, withsome wheat, barley and quinoa. Every village is within fiftymiles of Puno, Juliaca or the two or three other major markets.There were twelve extension agencies and fifty-three satelliteextension offices in 1969.

La Paz, Bolivia. The southeast shore of Lake Titicaca, with 600square miles, 190,000 persons and an altitude between 11, 500 and12, 500 feet. The study district includes the two provinces of LosAndes and Omasuyos, which belong to the department of La Paz,but does not include the city of La Paz or its immediate surround­ings. The culture is comparable to the one at Puno, with whichit was linked before the Spanish traced an artificial boundaryacross the Titicaca basin. There are two extension agencies inthe district.

Cochabamba, Bolivia. The Upper Valley between Arani andTarata, but excluding the potato regions in the hills on all sidesof the valley. This flat valley has an area of 200 square miles,with a rural population of 130, 000 and lies at about 8, 000 feet. Itis farmed for corn, wheat and some potatoes by mixed Indian andmestizo small holders. There are four agencies in the UpperValley including one that operates sporadically.

The "impact" study was performed in two phases. The first,the village level analysis, was made in seven districts using largescale maps and complete lists of villages. Districts with interest­ing experiences but inadequate map coverage could not be considered.An example of the standard 1/50, 000 scale map is reproduced here (map 2)to show the reader how clearly the villages show up. For each

~~.• l

Map 2

1: 50,000 SCALE SAMPLE

f.

TI

77

village I secured two sets of grades.

One set reflected the intensity of the village's contact withextension over a ten to twenty-year interval. For most districts,old daily logs were available which showed the villages visitedduring the agent's working hours. I counted visits by year for aperiod of about ten years, and assigned numerical grades,usually between 0 and 3, to each village based on the frequency ofvisits. To supplement this quantitative data base, present andprior agents were interviewed, and asked to assign grades toeach village to reflect the intensity with which they rememberedto have worked in each village. For the five districts where bothsources (logs and interviews) were used, the two types of mea­surements were aggregated according to a weighting process thatvaried from district to district depending on the author's hunchesas to the relative reliability of the grades. The author performedthese studies alone. He cannot offer independent corroborationof the grading results.

The second set of grades reflected the level of progress ofthe typical farmer of the village in 1969. These grades werebased entirely on subjective evaluations by selected observers,collected during an interview when the maps and village listswere laid out and the observer was requested to consider the levelof each village. Interviews were held with fertilizer dealers,credit officials, other non-extension personnel who were knownto have a familiarity with farming practices in a large number ofvillages, and, where these were not available, the extension per­sonnel themselves. There were usually two or more observersfor each village. Each observer graded the villages on a fixedscale, again, between 0 and 3, and the independent observationswere later weighted by the author according to their reliabilityand aggregated. As with the extension contact grades, the authorworked alone.

Using the two sets of grades, a matrix was developed show­ing the relation between the extension variable and the progressvariable: showing, for example, how many of those villages whichenjoyed the most contact with extension were considered veryprogressive, and how many of the very progressive villages werecharacterized by frequent extension contact. Over 1,500 villagesin the seven districts were graded. Table 3 shows some vitaldata for each district, and table 4 shows the results.

78

TABLE 3

VILLAGE LEVEL ANALYSIS: CHARACTERISTICS OF DISTRICT STUDIES

Approximate Villages Extension PeriodDistrict or Agencies Analyzed

District Study Population* Evaluation(Thousands) Units

t.

Graded •If

Quezaltenango 160 104 1 1954-68 ~

1Jutiapa - Jalapa 140 220 4 1955-68

Northwest El Salvador 550 373 21 1958-58 t.rt;

Boyaca 450 520 7 1954-60 r

54+~.

Puno 550 12 1962-68 I.,',,,

La Paz 170 175 2 1952-68

Upper Valley, Cochabamba 130 114 4 1954-68

* These are rough estimates. The Guatemalan and Salvadoran figures arederived from the list of population figures for only those villages includedin the analysis. The other figures are derived from census data at themunicipio or provincial levels.- Strictly speaking they are incommensurate,although the relative orders of magnitude are valid.

+ Districtos, which are groupings of villages.

; -

TABLE 4

VILLAGE LEVEL ANALYSIS: SUMMARY OF RESULTS

Percentage of all Villages Percentage of"Progressive"

Villageswhere the

Extension Service"Concentrated"

'(.55(4)

Percentage of"Concentration"

Villageswhich are

"Progressive"

where theExtension Service

I "never"Concentrat~d" arrived"

(2) (3)(1)

which are"Progressive"

!------.,r.-r----r..-----IDistrict Studies

Quezaltenango 67 21 57 100 31

Jutiapa - Jalapa 46 14 63 92 27

Northwest El Salvador 36 37 42 48 50

Boyaca

- village level- ~unicipiG level- Choconta*- Facatativa*

20 27 52 313325

"low"

3121

"low"" low"

+Puno 22 63 n.a. 32 92

La Paz 14 16 n.a. 41 46

Upper Valley. Cochabamba 27 16 25 73 45

Note: The'definition of "progressive" and "concentration" varies between districts. See text.

* Extension agencies in Cundinamarca.

80

The second phase of the "impace' study was performed atthe district level and carried on in seven study districts, includingtwo omitted from the village study for lack of map coverage. Thetwo Titicaca districts were dropped. Using information collectedfrom a wide variety of AID, host government and private sources,an historical profile of agricultural inputs and outputs, or indica­tors of inputs, was constructed. The amount and timing of theextension input in the district was plotted, and compared with theamount and timing of other inputs and economic phenomenaassumed to have been important (small-farmer credit, fertilizerprices, etc.), and with whatever indicator of material progresswas available. The indicator most widely used was the sale ofchemical fertilizer. The objective of the district level analysiswas to see whether extension or any other input had a clear anddominant relationship with measured progress. The village levelanalysis had not included the other input variables, and thus couldnot compare their role with that of extension. As examples of thegraphic relationships that were revealed by these district analyses,charts for EI Salvador and Colombia are repeated here (figures6 and 7). The first shows rather good correlation of extension aswell as credit with an increase in fertilizer sales in the 1960s.The second shows that the Boyaca pilot extension project inputslagged behind and probably do not explain the increase in fertilizersales that occurred during the 1950s.

A third phase was added to the productivity research pro­gram to complement the "impact" study. The addition was at thesuggestion of Donald L. Fiester, of the Latin American Bureau'sagricultural office, who felt it important not only to ask whetherextension contributed to progress in small-farm, basic crop zones,but also to look at the production programs where extension hasclearly made a contribution and to identify the characteristicsconducive to success. A series of "case studies" was thus added,most of them based on the sort of descriptive input-output analysisused at the district level. The "case studies" are not all successstories. They do not all feature the regular, national, extensionservice. But they all involve a development program that incorpo­rated a strong extension -type component, and the study design foreach was aimed at finding the ingredients of success or failure.

One important type of comparative analysis would have beento identify the types of extension techniques that were stressed ineach of the districts and cases investigated. The impression onegets is that all agents use all techniques (method demonstrations,

Figure 6

INRJTS AND FERTILIZER SALES IN THE NORTHWESTERN EL SALVADOR DISTRICT

2 200 40 20 10

1 100 20 10

"ABC CreditCountrywide

~ FertilizerSales, FiveTowns

Corn campaign impact,if any, would begin in1965

1965 1970

[lh1\

II,I \I \

\I \I \I \I \I

III

IIIII

[l

19601955

Agencies"""""

Cajas de Credito ----.Credit, Country-wide

Credit (Million Colones) - -G···.. ··

Corn Production (Thousand M. Tons)~Fertilizer Sales, (Thousand Quintales)~

Fertilizer Dealers~Extension Agencies •

fertilizer ~ealerstIncomplete)

5

60 30 15

80 40 20

300

400

3

4

Source: Miscellaneous material

Figure 7

Caja Fertilizer Sales,Zona de Tunja

~------ Caja Fertilizer Sales,Country\'ride

'----------------- Caja Agricultural CreditDeflated, Countrywide

G/ ...olf-------- Caja Agricultural

/ Credit, Current Prices,I CountrywideI

196019551950

20

40

60

80300 30

200 20

100 10

FERTILIZER SALES, CREDIT AND EXTENSION IN THE BOYACA DISTRICTShowing Caja Agraria sales and credit and

period of activity of the STACA extension project

Credit (Million Pesos)

IfFertilizer Sales, Zona de Tunja(Thousand M. Tons)Fertilizer Sales, Countrywide(Thousand M. Tons)

I400 40 100

Sources: Fertilizer Sales - Caja Agraria worksheets and notes retained by Gonzalo Isaza (Bogota).Credit - United Nations, El Uso de Fertilizantes en Colombia (New York: March, 1966), p. 15. Deflated withan i~licit GDP urice deflator provid~d by Ceotr~l,Bank (19~8=100)~ . r:9~ *STACA's impact on fertilizer salescould not have al'fected these sales flgures slgnlflcantly till l~5o or /7.

;-----•j

";_i

83

result demonstrations, field days, individual farmer contacts,group contacts, etc.), and the differences arise because of differ­ing emphasis. Unfortunately, information was not available tocarry out such a study.

Table 5 lists the field studies in the research program, in­cluding district studies and case studies. The nine study districtsare located on map 1 and their boundaries are shown in more de­tail on the maps in appendix a.

C. Findings

The nine major studies of rural districts within six countries,and the briefer survey of a number of agricultural developmentprojects in these and four other countries, fail to reveal any strongrelationship between the work of the national extension servicesand the cases of alleged progress in farm productivity.

In most of the seven districts surveyed at the village level,the list of villages identified as "progressive" in 1969 did notcorrelate well with the list of villages favored by the extensionservices during the last two decades. In fact, a Chi-square testshowed no statistically significant relationship in any of the sevendistricts. For the seven districts surveyed at the district level,available data on the timing and quantity of fertilizer sales, exten­sion inputs, credit availabilities, prices, market opportunities andproduction increases yielded no evidence that the extension ser­vices were instrumental in promoting improvements in farm tech­nology, and some clear evidence that other inputs were. Finally,the review of development case studies where extension serviceswere known to have played a role showed again that other factorsusually made more of a difference in determining the rate ofprogress than the extension service input. At best the extensionservices can be said on the basis of reputational and statisticalinformation to have played a minor part in the overall developmentof agriculture within the study area.

Those persons who would try to argue that field activity by anextension service was either necessary or sufficient for the diffu­sion of new technology in that area are wasting their time. Thereare more instances of progress which the extension service hadlittle or nothing to do with, revealed in the continental sweep ofthis stUdy, than there are cases of progress wherein the extension

84

TABLE 5

STUDY DISTRlcrS .o\ND CASE STUDIES

Showing classification

Village LevelAnalysis

Study Districts

District LevelAnalysis

Quezaltenango, GuatemalaJutiapa-Jalapa, Guatemala·Northwest EI SalvadorBoyaca, ColombiaAzuay-Canar, EcuadorMantaro Valley, PeruPuno, PeruLa Paz, BoliviaUpper Valley, Cochabamba, Bolivia

xXXX

XXX

XXXXXX

x

Regional Crop National Specialized PrivatePrograms-' Campaigns Extension Extension Sector

Service ServiceProjects Projects

Case Studies

STACA, Boyaca, Colombia XPlan LAC, Ecuador X X5 YP,Mantaro Valley, Peru X XSouthern Peru Program XSanta Cruz, Bolivia X XPlan Chillan, Chile X

Com Campaign, El Salvador X XPlan Masaya, Nicaragua X XCom Campaign, Costa Rica X XWheat Campaign, Bolivia X X

Honduras: Melons XHonduras: Black Beans & Corn XNicaragua: Cotton XCosta Rica: Coffee XEcuador: Bananas XEcuador: Chickens X

Puebla Project, Mexico XGreet.ii Guatemala XLa Maquina, Guatemala XCaja Agraria, Colombia XCoffee Federation, Colombia XCREA, Ecuador XANBE, Ecuador XSugar Industry, Chile X

FERTICA, Central America XTabacal~ras. Guatemala XUnited truif, Costa Rica X

85

service was clearly participating. Moreover, there are very fewexamples of extension service activities which appear to havetriggered a diffusion process: the extension service has notplayed the role of the "catalyst. "

It is possible to argue that the extension services have onmany occasions played the role of the "conditioner, " that is, thatthey have sensitized the farmers to a new system of technologyand opened the door to the development agents of those institutionswhich provided either the physical inputs in the technologicalpackage, or the credit to buy them. Extension agents maintainthat without the years of frustrating field work by t,hemselves andpredecessor agents, the farmers would not have been receptiveto the technological package when it was finally put together.More importantly, professionals of other institutions largelyagree. Research technicians, credit supervisors, fertilizer deal­ers and the like frequently stated that extension laid the ground­work for future success. If the extensio?- services have playedthis role, and if "conditioning" is an impprtant precondition ofprogress, then their contribution would hie more significant thanthe reputational and statistical evidence rrentioned above wouldindicate. That evidence would tend to favlpr the last inputs, andminimize the importance of the long cond~tioningprocess.

I

My feeling is that the "conditioner" tole is exaggerated.The extension services were rarely the only channels open tofarmers in the precondition period. and the other developmentagencies usually provided their own extension dUring the periodof growth. The work of the extension services in the preconditionperiod could have been important if it had been effectively ex­ploited: more often than not it was wasted. These are opinions.The study failed to turn up any conclusive evidence one way or theother on the conditioning role. We will return to this subject atthe end of this section.

Having deflated the extravagant claims about the role of ex­tension services, the study results show nevertheless that incases where the economic and institutional environments werepropitious. the extension services do appear to have played auseful part in speeding up the development process. No matterhow efficient the farmer-to-farmer channels of communicationare, professional expertise must be introduced into the system tobring the farmers to the optimal level of known technology andhold them there. The study does not prove this point either. but

86

everyone in the field said it was so and it seems too obvious tocontest. Thus the extension function is essential. If the exten­sion service is properly trained, and either integrated with therest of the development program in a planned environment, orcomplemented by other inputs in spontaneous growth, it can andhas managed to perform the extension function. In terms of someof the roles hypothesized for the extension services, we can saythat the findings of the study partly support, or at least are notinconsistent with the proposition that the extension services haveplayed the role of an lIaccelerator" in many (though not most) ofthe progress stories looked at. Its absence would have beenmissed but would not have precluded success. It played no rolewhen the environment was hostile to adoption of technology.

These conclusions seem inconsistent with the findings ofMichigan State University's new study of the rural diffusionprocess. A completion report of that three-continent study hasjust been made available to AID, which financed it. 15 EverettRogers, the study director, was struck by the importance of con­tacts with change agents as a factor explaining farme r and villageprogress in parts of Brazil, Nigeria and India. It correlatedmore highly with progress than any of the other explanatory vari­ables tested in the enormous MSU interview program. He says,in the abstracted summary, that "it seems that change agent con­tact is a necessary condition for diffusion in less developednations. It

In two countries - - Nige ria and India - - the national exten­sion service was the relevant change agency, and Rogers' argu­ment supports the extension agencies' traditional emphasis on in­creasing the number of contacts, and improving the communicationtechniques, as a strategy for success. In chapter 1 he evenannounces "the thesis of the present report is that the problem ofpersuading peasants to adopt innovations can profitably be ap­proached as a communication problem. II The ensuing discussionof important variables concentrates on sociological factors, andone looks in vain for an analysis of the effects of changes in theprofitability of an innovation on the rate or degree of adoption.

15Everett M. Rogers, )"oseph R. Ascroft and Niels G. Roling,Diffusion of Innovations in Brazil, Nigeria and India (East Lansing:Michigan State University, Department of Communications, 1970).

•

~ .,

,~l;

.~ .~f

87

The difference between the results of the MSU study andthis study could be attributed partly to the different methods forselecting villages. The MSU team only included villages in theanalysis which change agencies identified as ones where they hadassigned people to work. Thus, if there were progressive villageswith very little contact with the selected change agencies, the teamwould have missed them.

I suspect, however, that the MSU findings and this study'sfindings really do differ fundamentally, that Rogers' conclusionsoverreach his data base, and that his lIerrorll is probably attribut­able to unresolved problems of collinearity. Rogers was evalua­ting the extension function, not the extension institution, and thevillages and farmers the MSU teams talked with were enjoyingsimultaneous infusion of other inputs and favorable environmentalconditions. What we are probably observing in the monumentalendeavor by MSU is the desirable effects of including a changeagent in the action program, not the effects of dispatching exten­sion agents to every community.

If contact with change agents is not the decisive factor inrural development, what is? The next several paragraphs dis­cuss the roles of the other elements, in particular of credit, mar­kets, management, fertilizer supplies and research.

Evidence on the importance of credit and markets in explain­ing progress at the district level, and in the case studies of de­velopment, is overwhelming. I had anticipated before the studybegan that extension would correlate highly with progress, andthat arguments about the critical role of credit and markets couldbe shown to be exaggerated. I suspected that farmers couldscrounge from family and friends whatever money they neededfor a good investment, and that a surge in the production of acrop, even a dominant food crop, would within a few years andafter some worrying moments create its own market. Nothing ofthe sort. In district after district, in case after case, analysisindicated that pockets of progress in Central and South Americahave enjoyed either an infusion of credit or a favorable marketfor the local cash crop, or both, and that the absence of progressis associated more often than not with the absence of these factors.

The relationship between growth in fertilizer sales andcredit availabilities is particularly striking, as many of the dia­grams illustrated. One could predict on the basis of this study

-t-hat credit must be included in any diffusion program that is based

88

on purchased inputs. Perhaps this is not surprising. Creditenables farmers to buy fertilizer. Extension does not. Creditavailabilities would presumably correlate better with fertilizersales over time than would extension activities. The importantpoint, however, is that without credit extension services havebeen frustrated.

The market factor also appears to be decisive. If the rela­tionship between crop prices, yields and input prices is not ad­vantageous to the farmer, he would be unwise to adopt a newtechnology, particularly one that involved a heavy outlay on pur­chased inputs. The study revealed very few instances of pro­grams generated by the diffusion of costless or low cost technol­ogy. The argument that improvement in yields could be achievedsimply by reorganizing and rescheduling inputs and practicesalready available or in use on the farm (e. g., time of planting,density of seeding), and that these yield improvements were suf­ficiently large to encourage widespread adoption, is simply notsupported. The evidence suggests instead that significant im­provements over current yields will only be achieved throughintroduction of new inputs unavailable on the subsistence farm, atleast until the research establishment provides varieties muchmore resistant to pests and much more efficient in the use of rainwater and natural nutrients. Thus adoption has a price tag, andthe farmer will not buy unless the level of "expected" profits,after considering all elements of risk, was high. It was impos­sible in this study to determine the rate of return, or the guaran­tee against loss, that would induce farmer adoption. Studies ofthis sort are just getting going in the several centers of agricul­tural economics established in the study area. But it is possibleto say that most of the successful diffusion processes encounteredin the study depended upon a strong market and relatively stableprices.

There are enough interesting anomolies in the study tosuggest that another factor plays a role at least as important ascredit and markets. The wheat farmers in the Quezaltenangohighlands haven It been fertilizing their corn, even though the re­turn would seem to be no lower than the returns they get forfertilizing wheat. Yet the settlers in the La Maquina colony onthe South Coast of Guatemala have developed a most lucrativeurban market for their corn, and have reached a reasonably im­pressive level of corn technology. That market is accessible tothe Quezaltenango farmers. The virgin soil of La Maquina

,.

*,,.i

•:?'''f;

4

~1 :.s"~

'I

~";~

-'J:~

1!,1j1:t:~

~f,

};~';"I

i.;'I

'f./1,

jq.~

-l'-~'

~:~J;'I;j

;f);t04.Ii~;

\~

.t,-;~

";

;.?!,

~'i

:i'Ii"~

'iJ

i

\

'/'11:;"-

,1.,~,~,'t,~iJ.,}

89

requires less fertilizer than the overworked soils of the highlands.But this difference seems too small to account for the radicallydifferent rates in diffusion of corn technology. Similarly, thewheat farmers in the department of Bolivar in the Ecuadorianhighlands have begun to invest heavily in fertilizer, and theiryields are leaving the rest of the country's wheat farmers behind.Ecuador needs to import over fifty percent of its wheat; thus thereis a market for increased domestic production. Why did thefarmers in Bolivar move sooner than the rest? We also have thePuebla example, where CIMMYT appears to be successfully dif­fusing in one area a corn technology that could, with other varie­ties but little modification of basic principles or organization, beused in many different parts of Mexico. Etc., etc.

One finds many such cases where markets which appear un­attractive to one group have been exploited by another. What oneusually also finds in these instances is a higher level of manage­ment of the diffusion process in the progressive area. The man­agement is often provided by a producers guild or coop. It offersthe farmer help throughout the production and marketing cycles.Often it does more than help; it makes the important decisions forhim. The private tobacco and brewing company contracts withsmall farmers represent the fullest development of this manage­ment input. The Chilean sugar beet and Costa Rican coffee casesare other examples. Contracts are not always well advised, andin the long run may become an instrument of stagnation. Buttheir major advantage at the beginning of a diffusion process is thatthey offer the farmer security of enterprise. This seems to bethe most effective way to get rapid diffusion of an improved tech­nology. Some critics argue that the farmer under this systemnever really internalizes the modern practices and will revertwhen the contract is withdrawn. The farmer applies, but doesnot adopt, is one way to phrase it. I won't argue that matter,though it sounds phoney and doesn't test out in the examples fromGuatemala and Colombia I have briefly investigated. The impor­tant point is that the extension services have almost nowherewanted to assume the management function, and have usuallydeliberately avoided it.

Fertilizer availabilities also appear to make a differenceto fertilizer sales. This will surprise no one. Few small farmerswill travel long distances to buy bags or packages of fertilizer.For rapid diffusion of a fertilizer-based technology, a network ofrural outlets has to be provided. The gross temporal relationship

90

between the development of chemical fertilizer availabilities inLatin America, and the growth of progressive farming as re­flected by fertilizer sales, is no coincidence.

The input of research services into the progress storiesexamined here has been meager. Few of the diffusion processeswere based on the introduction of new varieties of familiar crops,or even the introduction of standard varieties of new crops (newto the area). The contrast between the research input into ruraldevelopment programs in the study area -- especially those pro­grams affecting small farmers who plant food crops -- and theresearch input into the miracle wheat and rice programs inAsia -- which the author has also studied -- is dramatic. ThePuebla Project again is an exception. It may be exactly what itssponsors imply it is, the first effort to bring research to bear onproblems facing small subsistence farmers in rainfed lands inLatin America. It is possible that the relatively low level ofprogress in the study area is attributable to the low level of re­search. This analysis would have no way of demonstrating thepoint, but it seems to be another good bet. An examination ofthis issue is essential.

Extension services have an impact on productivity whenthese other factors are present. They don't when they aren't.The remarkable and distressing thing about the extension serviceperformance is that throughout most of their history and in mostof the countries in the study area the extension services have nottried or been forced to coordinate with the institutions which pro­vide or control the other factors. Extension services have de­veloped an independent, informal. educational program, andpushed it vigorously whether or not conditions were appropriate.In fact, incredible as it seems. some extension services havepulled out of an area when the other institutions have entered it;some agents have disassociated themselves from a group offarmers if the latter were becoming clients of supervised creditor another organization; some directors have resisted letting theiragents be brought into a coordinated plan. Extension servicesseem in retrospect to have been seeking an environment in whichthey are least effective. This is not true everywhere. It ischanging rapidly now, often after the dismissal of the old exten­sion headquarters staff. But it is characteristic enough of the1950s and 1960s to explcHn almost entirely the very low correla­tion the study can demonstrate between the "map" of extensionservice activity and the "map" of progress.

r

'/

)!

,'I

j'\17

91

This is why the argument about the importance of the "con­ditioner" role is almost immaterial. It seems difficult to creditan organization that has a history of confronting the wrong prob­lems for a contribution that was at best ill conceived. This isalso why extension services ought to consider training theiragents to playa role which they are not familiar with, the one wecan call the "manager." It is a retreat, if you will, to the typeof job that could have evolved in the old fomento organizationswhich the extension services replaced. 16 It brings the agentscloser to the type of extension job which the special extensionservices in other public and private organizations provide, andwhich another study would probably show to have a much highercorrelation with progress.

One "danger" of assigning the extension function to a numberof specialized authorities is that the country ends up with a varietyof extension services. The extension school argues that this isbad, and it is easy to see how the system could suffer fromduplicated overhead costs and other inefficiencies. But the ad­vantages may outweigh the disadvantages.

I want to insert in the text at this point an important discus­sion on the redundancy issue by Robert A. White, written while hewas a graduate student in agricultural economics at Cornell.White feels the redundancy argument is exaggerated, and seescompelling evidence that the system which includes a variety ofextension operations may be more logical for Latin America thanthe universal service promoted by the U. S. technicians. The re­marks repeated here follow, in his text, a description of theColombian scene, where other commentators found far too manyextension -type se rvices.

"To the writer the conclusion that all the agenciesshould be joined into one national extension service is ahasty one and does not take into consideration that eachof these agencies came into existence to solve a specificproblem not being looked after by another agency. Eachhas developed a set of internal institutions which equipsit to attack certain problems of certain groups of farmerswhich are peculiar to the Latin American context. Much

16See appendix c-2. It is shown there that in at least fivecountries the U. S. supported extension services superseded older,promotional offices of the ministries of agriculture, offices whichwere usually (1) themselves the result of an earlier U. S. program

92

of the criticism of duplication is aimed at the semi­autonomous government agencies which have as a goalthe increase of production of a particular commodityor the private producer associations whicH are primarilyoriented to marketing. But these are the very agencieswhich are most 'indigenous I to Latin America and re­spond most closely to the economic and social values ofLatin America. They are also responsible for the rapidincreases in production and productivity which have beenaccomplished in Latin America. The problem is notprecisely the growth of many agencies. The very factthat they are spontaneous and are responsive to actualconditions rather than an import should indicate thathere are the potentially vigorous institutions whichshould be carefully cultivated and improved. The prob­lem lies in the coordination of the agencies that exist.If a coordinating, advisory body of agricultural agenciescould be established as was suggested by the IBRD andCrDA in Colombia, Ecuador, Peru and other countriesthe area of specific functions for each agency could bedemarked. Some agencies seem best fitted to workwith larger industrial farmers, others with the com­mercial family farms, still others with various problemsof the backward, subsistence farmers. "17

Before concluding this chapter, it is proper to ask whetherthe criticisms are fair. Do they apply to the general case, oronly to a few gross distortions? They would not apply to some ofthe servicio extension services during the early and mid 1950s,when agencies supplied fertilizer, pesticides and seeds and theextension role was understood by the best of the American andLatin agricultural officers to be only one part of a developmentproject. Ecuador during the Norman Ward period is an outstand­ing example. "Tiny" Ward was the U. S. Agricultural Officer andservicio director who promoted the poultry and banana diseasecontrol projects, highlighted in the general report. He was the"new" input which inspired Mosher to see better times ahead forthe Ecuador program in his 1953/54 survey.

and (2) offering an extension type service, at least part time.There was rivalry between these ministerial field staffs and theservicio extension staffs. The latter invariably won, and the otheroperation was dropped.

17Robert A. White, "An Evaluation of Agricultural ExtensionSystems in Latin America" (a paper presented to Prof. Donald

.,A

.;

93

They do not apply to Honduras, where the extension direc­tors fought successfully during the 1960s to preserve a fomentorole for their agents while losing most battles for recognition bythe "educational" school of extentionists at the international con­ferences on extension methodology. They do not apply to thepresent El Salvador extension service, under the leadership ofJose Perez Guerra. They do not apply to some of the agents inCentral and South America, those who on an individual base workclosely with research, credit, and fomento institutions and fertili­zer companies. They do not apply to U. S. agricultural officers ofthe vision of Milton J. Lobell and Albert L. Brown, who saw thatextension services were headed in the wrong direction and triedto force them back to play their appointed roles in a developmentprocess (making enemies for these officers along the way). Theydo not apply to the present wheat and wool extension program inBolivia, and to the practical advisory assistance provided to thoseprograms by the team from Utah State University. But they applyin the great majority of experiences investigated, and are fairlydirected at the general model, exceptions notwithstanding.

It is also reasonable to question whether the study designwas fair to the extension services, in the sense that the studyconcentrated on districts farmed by small holders and devotedto food crops, which circumstances were both less favorable toprogress. The vast majority of these small holders neithersolicited nor encouraged the growth of the extension system, sup­port which is said to have been critical to the development of ex­tension in the U. S. If the study had looked exclusively at the workof extension agents with tropical export crops - - like the experi­ences with small banana farmers in Ecuador and large cottongrowers in Nicaragua -- the conclusions would have been quitedifferent. But the study design can be defended, if only becausemost of the man years of extension work have been directed towardimprovements of the small farmers we selected. Moreover,specialist staffs have competed with extension agents in the ex­port crops while the extension agents have been almost alone inthe subsistence cultures, and that is presumably where theirprimary responsibility lies. Within the subsistence cultures, the

Freebairn for Agr. Eco 665, Cornell University; Ithaca: by theauthor, March 1968), pp. 35, 36 .

94

selection of study districts was certainly not unfair to the exten­sion services. As mentioned earlier, cases were selected wherethere had been some progress, and where the extension servicereportedly made special effort to promote it. This ought to havemade it easier to prove the case for extension. When I began thestudy I was convinced that extension played a decisive role andthat the results would support that conviction. It was astonishingto learn that progress achieved and extension efforts were largelyunrelated.

Finally, it is reasonable to ask whether extension servicescould have done the job we are suggesting they should have done.Would the other institutions have cooperated in a coordinated pro­gram, especially one under the direction of the extension staff?Were the other institutions -- credit authorities focused on smallfarmers, for example -- even in existence (in the 1950s)? Couldthe extension director have concentrated his staff in areas withfavorable economic environments, and thus further polarize thefarmers into two groups of haves and have -nots? (political forceswere pushing him in just the opposite direction, pushing him toopen offices in every district so as to expose farmers throughoutthe country to the magnanimity of the administration's rule.)Could the agents have managed a coordinated program, or identi­fied the economic opportunities actually available to farmers in agiven locale, when most of the agents had no more and often lessthan two years of formal preservice training, enjoyed almost noin-service training, knew little more than the farmers about farmtechnology and knew next to nothing about economics? Compound­ing all these difficulties was the budget constraint, which pre­sumably affected most government operations, but in the case ofthe extension services appears to have reduced them all to a com­mon denominator of absolute poverty that gave them no freedom toexperiment with plans that were not essential to implementing thefundamental extension model learned from the U. S. Given thesedifficulties, could the extension services have accomplished morethan they did? Probably not much more, at least during the late1950s and early 1960s. After, that is, the servicio phase -out andbefore the advent of the credit agencies.

One can, however, reasonably expect the extension servicesto have made more of an effort than they did to provide a usefulservice. The study shows that the extension function is effectivelyperformed only in favorable economic and institutional environ­ments. This interdependency was not missed by architects of theoverall U. S. technical assistance programs, and their Latin

"J..

95

counterparts, in the 1940s and early 1950s. But as the decadewore on and the servicio's were dismantled, extension servicescame under the influence of a school of thought that quite clearlypulled them away from the sort of accommodation that shouldhave been made. The U. S. technical assistance programs hadbuilt organizations whose role in rural development was uncertain,and which were not equipped after phase-out with the sort of admin­istrative flexibility and feedback that could have helped them bet­ter define a useful role. The Americans deserted them (though itis not certain whether a continuing American presence would havebeen beneficial since the Americans themselves seemed uncer-tain about the extension role). The institution building job re­sulted in a network of agents scattered all over Central and SouthAmerica who in retrospect appear to have been out of step withthe development process.

We can see that the logical separation of "effectiveness!!tests and "significancell tests is getting us into trouble, since itis difficult to call an institution builder effective if his institutionhas small significance, and it is difficult to fault the institutionfor having little significance if the institution builder should havebuilt it a different way. In chapter 6, some suggestions areoffered on the appropriate way now to salvage the extension ser­vices. The pace of change in Latin American extension is rapid,and already moving along the suggested line. I am unsure whetherby the time this study is read the Latin American readership willfind in it anything new.

CHAPTER 5

EFFICIENCY: COSTS, BENEFITS AND ALTERNATIVES

The tests of effectiveness and significance have not givengood results, an d the word failure has crept into the discussionof institution building and productivity. But the use of the wordfailure was premature, since it suggested there were better waysto accomplish AID objectives. That judgment, of course, dependsin turn upon a comparative analysis of the costs and benefits ofall the alternatives. Whether the apparently "unproductive" pro­gram in extension was a fiscal extravagance or a profitable ven­ture in technical assistance depends on the amount of U. S. andhost country resources that were tied up by it as well as on theapparent results. In fact, seen in this perspective, the net re­sults seem much better.

This part of the paper is more speculative than the pre­ceding parts. The information for a proper efficiency study wasnot available. In fact it would call for a full analysis of alternativestrategies for rural development, which was beyond the scope ofthe paper. But the elements of the efficiency study must beexposed. An evaluation study that fails to do this, however in­teresting, cannot be used as a tool for improving decisions onforeign aid and rural development policy.

A. Costs

Table 6 shows the cumulative costs to the U. S. governmentof the extension programs in the study area during the institution­building period. The period starts in 1942. It ends in the mid­1960s, though the year of completion differs from country tocountry. It does not include the modern investments in extensiondescribed in chapter 2.

It was not possible to retrieve accurate data on capital andtechnical inputs. Servicio accounts are in many instances notavailable, and even if they were there is no assurance that abudget line item giving the extension outlay includes the samecomponents (the agricultural information service, training,vehicle repair, etc.) in all countries. Moreover, some countriesreceived extraordinary capital grants and PL 480 proceeds, and

97

98

TABLE 6

COSTS'-1'0' TIlE U.S. GOVERNMENT OF THE EXTENSION PROGRAMS

Expenditures

Thousand of Dollars

Country Year Local Costs Dollar Costs

OtherJoint Budget U.S.Funds Support Technicians Other Total

Guatemala 1955-64 660 1,040 n.a. 400 2,100El Salvador 1948-62 0 100 It 400 500Honduras 1951-65 950 80 It 270 1,300Nicaragua 1950-58 350 0 It 150 500Costa Rica 1942-65 600 100 It 400 1,100Panama 1942-62 0 100 It 100 200Colombia 1954-62 500 300 It 300 1,100

Ecuador 1955-64 1,700 200 " 300 2,200

Peru 1943-62 760 400 It 340 1,500Bolivia 1948-66 1,320 1,730 It 450 3,500Chile 1954-62 750 0 " 350 1,100Paraguay 1950-69 550 300 It 250 1,100Miscellaneous* 1,800 1,800

Total 1942-69 8,140 6..1.50 12,000+ 3,710 30,000

(Percentageof Total) (27) (21) (40) (12) (100)

Source: Miscellaneous material.

Note: All figures are rough estimates. The ones for Nicaragua and Panama arebest guesses, since primary data was unavailable for these two countries.

*Figure introduced to cover unattributable local costs of extensionprojects funded from PL480 and other local currency accounts in Guatemala, Bolivia,Chile, etc.

+Seven hundred seventy five man years, at $15,000 per year, amounts toapproximately twelve million dollars.

:J

99

it is often impossible to find how much of these funds were spenton extension (for building extension offices, for purchase ofdomestic seed, etc.).

The biggest problem was in fixing the imputed costs forsubject matter specialists who were not full-time extensionists.It is assumed in this table that all such technicians gave halftheir time to improving the extension service, and thereforehalf of their costs to AID have been attributed to extension. Thisundoubtedly overstates the contribution, but that in turn is com­pensated by the practice in this table of not including the time ofthe chief officers in the agriculture mission. U. S. contributionsto the overhead costs incurred by other offices but attributableto the extension operation are also excluded. I will not spell outall of the assumptions upon which the figures in the table arebased. Suffice it to say that they are very rough estimates, andprovide only an order of magnitude of the country and area totals,and the relative magnitudes of country and component shares.

The total cost to the U. S. gove rnment for building extensionservices in the Andes, excluding the overhead costs, is approxi­mately $30, 000, 000 (in current dollars). Bolivia got the biggestshare, because of the sustained budget support operation. EISalvador and Panama got the smallest shares, because they re­ceived almost no budget support. Of the total, about 480/0 of thedollars went to budget support for local costs, 40% went to thedollar costs of U. S. technicians, and 12% went for U. S. commodi­ties, participants and other dollar costs. If we pretend that the$30,000,000 was all spent during the fifteen year, major institu­tion building period between 1948 and 1962 (as most of it was),then the armual outlay amounted to $2,000,000 and the averageannual outlay for each of the twelve countries was $166,000.These figures are surprisingly small. An adjustment to bringthem to constant 1970 dollars or to add concealed overhead chargeswould not alte r that conclusion.

The costs to the host governments were computed indirectly.

Table 7 shows the annual cost per agency of the extensionservice budget in each country in some recent year. Thus, forGuatemala, the reported 1968 extension outlay of $233,000 wasdivided by 37 (the number of agencies), giving $6, 300 per agency,the figure in the table. Bolivia's relative poverty is apparent.So is Colombia's new state of affluence, and the relatively highcosts of agencies in Costa Rica and Peru (Nicaragua's apparent

100

TABLE 7

COST OF TIlE EXTENSION SERVICES, PER AGENCY

Extension Budget

Country Year Agencies Local Foreign Dollars CostsCurrency Exchange Per Agency

Rate(Thousands) (L. C./Dollar) (Thousands) (Dollars)

Guatemala 1968 37 233 1 233 6,300E1 Salvador 1967 49 854 2.5 '342 6,980

Honduras 1968 33 510 2 255 7,730

Nicaragua 1968 21 2,363 7 338 16,000

Costa Rica 1968 38 3,395 7 485 12,760Panama 1963 20 341 1 341 17,050Colombia 1967 24 3,200 16 200 8,330

" 1969 50 29,000 17 1,737 34,740

Ecuador 1968 41 5,265 22 240 5,850

Peru 1968 141 84,300 44 1,920 13,620

Bolivia 1966 65 3,400 12 272 4,185Paraguay 1968 22 26,500 126 210 9,550

,'; ~

" 1969 24 17,500 126 139 5,800

Source: Miscellaneous material.

Note. Budget figures usually show expenditures, or estimates of expenditures.The Ecuador and Bolivia fiaures are probably inaccurate. FX rates are approximate

and based on the International Monetary Fund's International Financial Statistics,XXIII, No.1 (January, 1970).

101

ranking as one of the most expensive services does not conformto material evidence of prosperity in that service, but I am at aloss to explain the discrepancy.) The costs per agency will varydepending on how well it is equipped and how big a staff it holds.The high ranking of Peru, for example, reflects in part the costsof the junior agents -- the so-called "sectoristas" -- who man theagency I S satellite stations.

The table is useful for providing an order of magnitude ofthe average annual cost of operating an agency. The averagecosts are undoubtedly lower now than they were in the serviciodays, when the budget was more generous. On the other hand, in­flation has eroded the value of the budget, so that a $6,300 agencyin 1955 was better equipped than a $6, 300 agency of today. Con­sidering these opposing trends, and the tabulated figures, it seemsappropriate to select $10,000 as an average estimate of the annualcost in current dollars of running an agency in the Andes since1942. The typical agency thus identified would have an agentwithout a college degree, two assistant agents, one-half of a homeeconomist, a male secretary and one and one -half jeeps. Rentwould be supplied free by the municipal authorities.

The number of agencies operating in the study area each yearthrough 1969 was shown in figure 5. Altogether, for ten of thetwelve countries (Panama and Chile excluded), there were 6, 155agency years. Round the figure up to 7, 000 to include Panama andChile. At $10,000 per year the total bill amounts to $70,000, 000.Of this, we have said the U. S. Government covered $15,000,000through budget support. That leaves a bill for the host govern­ments for the twenty-eight year period of about $55,000,000. Thiscan be added to the U. S. contribution of $30,000,000 for a totaloutlay of $85,000,000. Administrative and other overhead chargesto other offices are again excluded.

B. Benefits

1. Institution Building

If the extension organizations that have been developed wereand would continue to be worthless, then any investments in theinstitution building process would have been wasted. They are notworthless, however. They have provided a training ground forprofessional agriculturalists in the public and private sectors.

102

They have partially converted government bureaucracies to a newappreCiation of the farmer and of their responsibilities to him.They have improved the quality of life for many communities,farm wives, and rural youth. And they have contributed in asmall way to the raising of farm productivity and cash income.Moreover, they form a substantial organizational infrastructurethat links governments to farmers all the way from Guatemala toChile, an infrastructure that could form the backbone of adynamic rural development program. Al though the study wasunable to quantify any of these effects, it seems plausible toargue that the sum total represents no small achievement. Thefact that the U. S. foreign aid administration "bought" it for$30,000,000 does not immediately suggest fiscal extravagance.Whether the same sum invested in a program loan, a hydro­electric facility, of land reform would offer significantly greatervalue is not at all certain. I cannot pursue this line of reasoningany further, however, since it demands comparative data that isnot at hand. The fact that the extension services were not usedas well as they could have been is another issue, and will be dis­cussed separately.

2. Increasing Productivity

The cash value of the productivity impact of extension couldnot be empirically determined. However, it proved useful to con­trive a simple model of the agent-farmer relationship. This sug­gested the relative magnitudes of actual benefits and costs peragency, and the variables that mattered most to the agent if hewere to have a large impact on productivity.

The model was based on an assumption that the target in­crease in income per farm was $200 per year. The model thenindicated that the agent had to convert fifty farmers each year tobreak even (50 x $200 =$10,000). This approach assumes thateither the farmer reverts the next year, or that income in futureyears attributable to the conversion is not credited to the agent.Both assumptions are unrealistic. Moreover, if farmer to farmerchannels were active, and the multiplier factor operating, thenthe agent need convert less than fifty, since his converts would

-~~,rfIi'J

I .

iJ

i'I

),1

".~JJ~

,~

~

1i~ .1:rJg

103

convert the rest. But let us ignore these adjustments, 18 andsimply ask whether estimates of annual individual conversionsranging up to fifty are realistic.

Let us assume that to produce the conversion we are talk­ing about, the agent has to visit the farmer, or the farm group,three hours a day, every other month, for one year, and that'sall. We will say that that amount of contact doesn't guaranteeconversion, but that it is necessary for conversion. The agentcan make three such visits per day. With a five day work week,and four weeks' vacation, the agent thus makes seven hundredtwenty visits per year (3 x 5 x 48), or six visits annually foreach of one hundred twenty farmers or farm groups. If conver­sions run at the rate of 500/0 of contacts, then we could imaginethat in this system the agent (working together with his staff)could expect to make sixty conversions each year. The enormousleverage on this figure of the extension multiplier, if it operates,should be obvious. In fact the model with the multiplier predicteda veritable army of converts over a ten-year period and sug­gested an impact on incomes that was quite impressive.

Too impressive. The sort of production/income impactsuggested by the model seemed way out of line with observedresults, so instead of using the model for the purpose for whichit was designed -- to estimate the level of benefits attributableto the agency - - I assumed that the level of benefits was low andworked the other way through the model to see what was wrong withit. The obvious conclusions are that the number of permanentdirect converts is much less than fifty, and the multiplier has notbeen operating. This would be the case if (1) the new technology'soffer of profit was neither substantial nor reasonably certain, and(2) the farmers knew it. The real utility of the model then is tosuggest what could happen if economic incentives were provided,and especially if the agent switched from working with individualsto working with groups.

18"These adjustments" are the only thing that could give themodel legitimacy. What is needed is to discount all future in­comes to the agent's converts, and to the converts' converts,attributable to the agent's work in any given year, and then allowfor a "steady state" of continuing extension inputs year after year,to pick up any reinforcing (or cancelling) effects one year's con­versions may have on the next year's conversions.

104

Having diverted the model for another purpose, I am leftwithout a reliable method for estimating benefits. This is thecentral weakness of the study: its inability to quantify the amountof farm income, real and imputed, attributable to the efforts ofthe average agent. The study results show that the cumulativeeffects of extension activity in twelve countries has not made amajor contribution to whatever progress has occurred, and thatwhenever the density of agents has been increased for specialextension projects, but without prior, reinforcing changes inthe economic environment, the situation was not improved. Theresults suggest that the benefit/cost ratio is low. But the ratiomay be higher than one, and it may be higher than that of someother rural development inputs. The only things that can be saidwith any confidence are that the U. S. effort, in cooperation withthe host countries, to transform the rural economy via the con­ventional extension route has failed, and that routes which wouldhave used the extension service in a different way, and whichalmost certainly would have offered better results, were ignored .

. , /' ,.But something else must be said to balance the discussion.

Thus far the spotlight has been on the average agent. The reallygood agents are a separate breed. Invariably they assemble theother inputs, and mix coerCion and persuasion with friendlyadvice in ways that are barely detectable. The most instructiveand rewarding moments of my field trips were spent in the com­pany of such men, and of the farmers who beseiged them. Theseagents ought to be studied apart from the others, and assigned aseparate benefit/cost ratio. The latter might easily exceed that.'attributable to any input of equal expense that the go~ernment

could assign to the rural development program. If good agentscould be recruited and retained, the need to seek other ways toput extension to work would be less urgent. But these agentscan be neither recruited nor retained.

C . Alternatives

This section asks whether what the U. S. technical assist­ance program accomplished, or hoped to accomplish, could havebeen accomplished more efficiently. For expositional purposesthe issues are raised as a series of three questions.

I!,

'.''} .

105

Could the same, conventional, extension organizations havebeen built more efficiently with a different technical assistanceprogram? This was to have been a major concern of the paper,to see if preferred strategies for delivering such aid could bediscerned. A few tentative conclusions were drawn. For example,there is evidence that a higher ratio of subject matter specialiststo extension methods specialists would have improved somecountry programs. Also, the practice of building extension ser­vices within the servicio was challenged, on the grounds that itinhibited the assimilation of the program and staff after transferof the service to the ministry. The vigor of EI Salvador's exten­sion service, which remained outside the servicio, was cited.But there were no strong signals from the comparative analysison many other crucial matters -- except, that is, on whether toprovide budget support. This seems, at certain levels at least,to be counterproductive.

The fact that no preferred strategy emerges does not meanthat the design of technical assistance programs doesn It matter.Rather, it reflects the post-transfer poverty of extension agenciesthroughout the study area, poverty which has wiped out the earlierand possible potential successes and left all U. S. programs look­ing, in retrospect, rather ineffectual. So it appears that budgetsupport is bad, and the consequence of failing to provide budgetsupport is also bad. I do not want to pursue this subject, sinceit is doubtful whether the Jl same, conventional, extension organi­zations ll should have been built at all. That would explain why,when they were built, nothing seemed to work very well.

Could the extension function have been organized differentlyso as to have a greater impact on agricultural productivity? Theaverage extension agent had a low impact on productivity because(1) the conventional system persuaded him to busy himself halfthe time with unproductive activities; (2) the conventional systemdid not develop his technical competence, so that he usually hadlittle to offer when talking about production; and (3) the conven­tional system did not encourage him to link up with supplementaryservices, so that on projects where he could offer a profitablenew technology, many farmers either couldn't find the requisitesor couldn It buy them.

In chapter 4, a few alternative institutional arrangementswere hinted at. Such a list would include,

106

(1) Extension units within single-crop production andmarketing authorities;

(2) Ministerial extension services authorized to coordinateother program inputs in crop campaigns, regional develop­ment projects, etc.;

(3) Supervised or directed credit operations, based onintensive farmer-agent contact;

(4) Land reform agencies, equipped with extension services,credit, and the power to intervene to break the large land­holders' control of crop markets;

(5) Research! extension services, where the agent isinextricably linked to a technological base and the emphasisin research is on profitable crops.

(6) Village-level-worker systems, as designed for theRepublic of India, provided that the subject matter special­ists and resource agents really do respond to the VLWagents I requests.

The list is much longer since variations and mixtures areendless. Most of the examples would have been written off by the"extension school" as fomento, since they involve prearrangedpackages of inputs, specification of suitable crops, and economicincentives. One alternative to the present pattern frequentlysuggested - - rapid build -up of the subject matter specialist staffwithin the existing extension se rvices - - would not in itself sufficesince it offers no means to overcome the single most importantconstraint on extension impact: the lack of economic incentive.

The advantages of the proposed alternatives are chiefly dueto the incentive package, and the strengthened linkages betweenthe extension agents and the other change agents: they ensurethat the extension function is organized as a tool of a developmentprogram, rather than as an alternative development program.Translated into the language of section b of the present chapter,the argument would be stated in the following manner.

The dynamics of imitation, which appear as secondary andsubsequent conversions, is a fundamental element of successfulsmall holder development programs. The diffusion agent ought

/j

•

107

to serve primarily as a catalyst, though he will serve later tokeep adopters informed and hold them at optimal productivitylevels. No diffusion agency can contact all the farmers, thoughgrouping them greatly expands the number of contacts. Groupsshould not get so big as to dilute the message. On the other hand,large groups of farmers can reduce the social and psychologicaldisincentives to change for each member of the group. Theseand other inter-farmer communication channels must be exploited.Farmers must be induced to imitate their neighbors, which meansthe technology must be profitable and relatively riskless, and therequisites must be available.

The conventional extension operation offers none of this andelicits practically no secondary conversions. Some agents arenot even interested in secondary conversions. When I asked theextension supervisor for the western region of Guatemala how hewould assess the impact of extension in his area over the lastfifteen years, he said he would go to the original cooperatingfarmers and see which improved practices they had retained~

The integrated programs, on the other hand, and presumablythese can be spearheaded by extension-type officers, succeedbecause they do induce spontaneous conversions at a great rate,and because they can keep secondary as well as primary convertsin business. In fact, that might be the definition of a successfulrural program.

Could the productivity effects attributable to the conventionalextension services have been achieved more efficiently under adifferent foreign aid program? If the objective was to increaserural farm incomes in the study area by the amount the extensionorganizations have increased them, should the U. S. have investedits $30,000,000, and the host countries contributed their$60,000,000, to land reform, supervised credit, plant breeding,livestock improvement or some other rural project instead ofextension? We don't know. If you ask "could the same resultshave been achieved more cheaply, or more qUickly?" The answeris probably yes. But if you ask "would the same results ... ,"the answer is probably no. AID has not distinguished itself for itsability to identify optimal rural development strategies, and ananalysis of $ 30, 000, 000 rural programs in other regions wouldprobably show some that did better and some that did worse.

We see much of this only with hindsight. In 1950, when theextension machinery began to roll, it is not clear whether theapparent alternatives appeared to offer much more. It is

---- -------------

108

necessary, therefore, to repeat an earlier remark, which wasthat the $30,000,000 and $60,000,000 bills already invested isnot necessarily too high a price to pay for the organizationalstructure and the goodwill that has been generated, provided theyare now properly exploited.

I1!

CHAPTER 6

CONCLU SIONS AND PRESCRIPTIONS

A. Conclusions

U. S. efforts to build agricultural extension institutions intwelve countries of Central and South America achieved neitherthe institutional security nor the production impact that were in­cluded among the original U. S. goals. For this reason, for otherreasons having to do with misconceptions about the role of exten­sion in rural development, and for factors related to unsuspectedfaults in the servicio device used to administer the technicalassistance program, the institutions the U. S. tried to build havenot matured properly and are now under attack by action-orientedprofessionals everywhere. On two of the criteria by which weare evaluating the U.S. program -- effectiveness (whether theU. S. built viable institutions) and significance (whether the institu­tions increased agricultural productivity) -- it rates rather low.Some of the most important findings of the study which supportthis summary judgment are:

1. Extension organizations were usually developed withina servicio and achieved there a high level of activity andrecognition. But they were not prepared for transfer tothe bureaucratic environment of the host government.They were met in some countries with unconcealed hostility.

2. A certain arrogance had already developed among exten­sionists, and this factor, plus an unfortunate tendency tounderrate the importance to the extension mission of func­tional linkages with other institutions, helped keep exten­sionists isolated after transfer.

3. The extension operation proposed for Latin America isexpensive. But budget officials in almost all countriesremain unimpressed with the extensionist's claim to alarger share of public funds. In fact extension has beenstarved ever since transfer, and this partly explains thepoor results.

109

110

4. No solution was devised for the problem of low budgetallocations. The servicios had provided good salaries, andfunds to cover extension's material expenses. The minis­tries could not promise either, even when they wanted to,and usually were forced to set low salary scales for theagents and eliminate most of the other items. While thissituation was not unusual for government services, itrarely crippled those services as much as it did extension.The problem could have been anticipated. With some fore­sight, the Americans would have realized that the conven­tional model could not be sustained after transfer. A countof the number of farmers that ought to be involved in theextension programs, and of the number of farmers anyoneagent could handle, would have indicated an extension staffof a size that neither the local budget nor the local trainingfacilities would permit. The Americans and their counter­parts ought to have begun very early to experiment withgroup contacts, integrated programs, and other techniquesto multiply the impact of the scarce agents. They should al­so have questioned the advisability of designing extensionprograms, and training a generation of agents, to dependupon a well maintained fleet of four-wheel drive vehicles.

5. Extension operations deserve a high rating on somecriteria of successful institution-building, and low on others.The image of extension, that is, what people think otherpeople think about extension, is remarkably better than theestimated impact on production suggests it ought to be.

6. The IB model does not tell us how to evaluate anorganization which is weak in certain attributes and strongin othe rs. In the cas e of extension, however, it wouldappear there is no possibility of tradeoff of linkages, andthe fact that the extension services in the study area failthe linkage tests in most important respects would appearto be sufficient grounds for failing the entire institutionbuilding process, the commendable achievements inother institutional tests notwithstanding.

7. This study has paid particular attention to the impact ofextension on agricultural productivity. The subject hasintrinsic interest to a development economist, as well asoffering another indicator of the success of the institution­building experience. The results of nine district studies

j

111

and a larger number of case studies of individual extensionprojects indicate that extension has played a marginal rolein most of the improvement in productivity that has occurredin the study area. The extension services have been neithernecessary nor sufficient as an instrument of development,and there are more instances where extension almost cer­tainly had no role in a diffusion process than there areinstances where extension may have had a role.

8. Extension services appear to perform the essential ex­tension function only when the services have linked spon­taneously or by design to research, supplies, credit, andmarketing authorities in profitable production programs.

9. Failure of extension to strengthen these linkages andthereby help raise productivity levels will probably leadplanners to ignore them or destroy them.

10. It is already questionable whether the extensionorganizations as presently constituted will survive. Somehave been reorganized and reoriented. Others will followthe same route. Because of the institutional and pro­grammative image it creates, the word extension is evenunder attack.

11. The mistake may have been made at the very beginningof the institution-building program. The extension opera­tion should not have been institutionalized as an independententity. This is because the only place an independently in­stituted extension team could play its proper role was eitherwithin the servicio environment, or in a ministry where thedirector of agriculture insisted on the closest coordinationbetween his units. Neither of these conditions obtainedpermanently, or customarily, in the study area. Thus thea ttempt to build and transfer to the host government theindependent extension institution was ill-conceived. 19

19Institutionalization of the extension function in the U. S.worked, it is said, because the other factor inputs were widelyavailable and because certain organizations of institutions -- suchas the land grant college -- set boundaries to the extension school'sgrowing ambition to remake society all alone. We do not say thatthe extension function should not be institutionalized, only that in­stitutionalization was a mistake under the conditions obtaining inthe study area, and in U. S. aid relations, during the study period.

112

12. The independence and isolation of extension led tostrange role assignments in both U. S. and Latin sponsoredrural development programs in Latin America. Extensionadvisors were put in charge of some production programs,and extension teams were given areas of the country fortheir exclusive attention. But the advisors and the teamswere trained in methods, not in content, and in any casethey couldn't function alone. This is like putting a camera­team in charge of a movie production. It can work with aspecial breed of camera technicians. The U. S. extensionadvisors and their counterparts were not of that caliber:their training was not suitable and the recruitment processselected them on criteria other than ability to design andcarry-out integrated rural development programs. Theywere, in effect, out of their element and some of them arethe first to admit it .. The extension teams in special,extension oriented, area development programs, like theill-fated one in Boyaca, Colombia, were on an impossiblemission - - without the requisite gadgetry.

13. The proper place for the extension advisor is on thestaff of a fomento institution, teaching diffusion agents theprinciples of communication, motivation and farmerpsychology. The proper place for extension agents, unlessthey are well trained in subject matter, is in the fieldservices division of a fomento institution, where they areinextricably linked to the rest of the development team.

14. The leadership and the agents in the extension servicesof the study area, though undertrained for the job at hand,are in general very good men. But because of an ill­conceived conceptual inheritance, they are a wastedresource.

15. U. S. technical assistance agencies must accept respon­sibility for the way the program has developed. The U. S.got into the building of extension institutions for the wrongreasons and got out of it for the wrong reasons. It is toAID's advantage to help bring extension back to life, inorder that it playa pivotal role in development.

,1ij

~

.~

~,II;,

,t

1 i

,2

'1I

1j1~

',1,1"

i,,\,t,~

JfJ.•~;:~

IfIi,

'~;j

);f~J

113

Such criticism is warranted, but leaves out much of thestory. There is a theory that says that the process of institutionbuilding by outsiders will consist in the building of some sort ofa structure, its apparent collapse when the foreigners leave,and the emergence of something indigenous out of the ruins andexperience. 20 Thus present difficulties were expected, the testof success of the American venture remains for the future, andany evaluation at this point is premature. The theory is attrac­tive' but in this case may be overly generous. It is more con­vincing to argue in support of the extension program that althoughthe particular extension institutions the U. S. tried to build havedone poorly, the concept of extending technical information fromsources to users has been widely accepted. The proliferation ofextension capabilities in rural development agencies must bepartly attributed to the U. S. program, and to the agriculturalprofessionals who were trained within it. Moreover, there areother significance effects of the program which must be considered,such as improvements in the stature within rural society of women,youth, home affairs and community affairs. The sum of theseeffects, including the goodwill generated by the extension pres­ence in the countryside, is considerable.

The reader may not agree that the benefits outweigh therather modest costs of the program. These were estimated at$30, 000, 000 for the U. S. Government, and $55,000, 000 for thetwelve host country governments, for the 28-year period 1942­1967. But the reader will have to accept that the number of agentsfielded under the program is a fraction of the number recommendedfor the conventional extension job in the study area. FAO and theFES recommend a ratio of one agent for every 500 farm families.The ratio in the study area is currently one agent for about 10, 000farm families. The extension lobby can easily argue that itssystem has never been given a fair chance.

I remember a comment by Jorge Ramsay, one of the elderstatesmen of Latin American extensionists and a man who got histraining in Chile rather than the U. S. He was discussing the gapbetween expectations and resources, and comparing extensionwith the formal education system. He noted there were 30, 000school teachers in rural areas, and still a considerable amount ofilliteracy. He felt it unfair for the budgeteers to "give him 30 ex­tension agents and expect him to change the world. "

20Conversation with J .K. McDermott, AID/TA/RUR.

114

Is the solution simply to increase the number of agencies andgive the agents good transport? That is what the Nathan Associatesteam seems to recommend for EI Salvador, in a 1969 report en­titled Agricultural Sectoral Analysis for EI Salvador, prepared forthe host government and for USAID. 21 The Nathan team is notunaware of the importance of better training for the agents, ofsubstantial investments in agricultural research, and of favorablemarkets. In fact the emphasis in the report is on price phenomena.But the role for extension in the sector strategy is spelled out inconventional terms, and the resources recommended for extensionare measured by the number of farmers in the country and theratio of one to five hundred.

Such mechanical projection of extension needs may work inthe U. S., where market data is widely shared, farmers seek tech­nical information, and agents have it or can get it quickly. Inthese circumstances the appropriate number of agents may indeedbe deduced from formulas which are based on the estimated aver­age time per telephone call and the estimated average driving timeto a farm. Five hundred farmers may be right, given a five daywork week and four weeks of leave and sickness. But in LatinAmerica, where markets are poorly developed, farmers are sus­picious, the agents are under-trained and haven't much useful in­formation to extend anyway, ratios of any sort are almostirrelevant. The ministry of finance is not going to fund the recom­mended level of agents, especially when it is based on contactswith individual farmers. The ministry isn It going to fund any­where near the recommended level. Recognizing this, one has tothink about priorities -- production priorities, regional priorities,employment priorities. One has to plan to take maximum advan­tage of group dynamics and the powerful multipliers that canoperate on extension contacts. One has to crowd agents into pro­grams where incentives obtain naturally or by public decree. Andone has to assess the extension job in relation to the other inputsrequired by integrated programming. The manner of determiningextension needs in this resource-scarce system is completely dif­ferent than the system suggested by Nathan Associates.

Contract teams of the Nathan Associates caliber are not un­aware of the dangers of simplistic formulae. But one wonders if

21Robert R. Nathan Associates, Inc., Vol. III, pp. 108-110,and Vol. IV (Summary), p. 88 S.

115

they can anticipate the effects of their own recommendations onAID programs. It is likely that the Nathan Associates' recom­mendations, if accepted as a basis for designing a technical assist­ance program in agriculture for the 1970s, would result in thestrengthening of the USDA PASA team in EI Salvador. Judging fromexperience in other countries, it is then possible that that teamwould consist of an extension specialist who sits at Santa Tecla(until the proposed CENTA is built) and seeks program guidancefrom the FES in Washington, several research specialists who sitat Santa Tecla and refer to FAS and BPI in Beltsville, Maryland,and a marketing specialist who sits in San Salvador and reports tothe ERS in Washington. The U. S. extension specialist might notbe properly integrated with the rest of the team, 22 his advice tothe host extension staff might not relate intimately to the otherUSDA advice, and the extension program that materializes mighttend to be independent of the national strategy. What could savethe situation is the unusual, incumbent director of extension, JosePerez Guerra, who would s.ort out the advice that's useful fromthe advice that isn't, and keep extension as much as possible inthe forefront of rural development action programs.

The future goals and program for extension must be writtenby authorities higher than extension. There must be more beanbreeders, if you will, calling the shots. If it is asked "what doesa bean b reeder know about extension?", the answer is that he knowsmore about extension than an extensionist knows about bean breed­ing, and that answer will do.

Something went wrong in the U. S. extension movement whenit tried to recreate an action system overseas. As suggestedearlier, the professional extension school's efforts to codify theextension doctrine led to the design of a "conventional" model ofan extension service that bore little resemblance to the "historical"U. S. extension experience. The school seems to have gotten overlyinvolved with a particular and ill-conceived methodology of informaleducation, and lost sight of production goals that weren't beingachieved. The fundamental conceptual error was to foreswear theuse of manipulative techniques and the substitution of unfelt needsfor felt needs. True, the conventional system would have done

220ne USDA extension adviser working in an AID program inSouth America told me that as far as he was concerned his only bosswas the FES's Assistant Administrator for International Extension,despite the fact the latter individual has tried to discourage suchindividual allegiance.

116

much better had adequate market incentives existed. But the exten­sion school refused to admit that the lack of incentives was holdingback progress, or that extension without incentives was useless.The U. S. advisers and their counterparts tried to institutionalizethe extension function in an autonomous institution that had "its ownthing" to do, and in the process cut loose from an historicalAmerican model which would have served them better.

The best example of this spirit of independence was presentedin my last visit to a study country. It was Paraguay, and it refersto the ex-director of the Paraguayan extension service. He hadcome up from the ranks during the long period when leadership wasin the hands of an American servicio officer. He was appointed tobe the first director of local origin when leadership was transferredin 1966. He is a very personable fellow, and dedicated to the idealsof extension. But he had been isolated and eventually removed frompower by production oriented activists elsewhere in the ministryof agriculture (activists, incidentally, who were trained in the U. S.in the plant sciences). They had grown progressively impatientwith his refusal to cooperate with other government units in produc­tion programs. He had dragged his feet over getting his agentsinvolved in the wheat campaign. He refused to have his agents getinvolved in a new rural credit program. He didn It attend meetingswhere national priorities in agriculture were discussed.

He was convinced that they were wrong and that he was right.He said he had one of the best 4-H programs in South America, andthat in itself was an achievement. He felt the purposes of extensionwere being distorted, and he wondered aloud why AID, which repre­sented the country which really "understood" extension, had lethim be outmaneuvered. The conversation ended with a commentwhich I believe captures the glory and the absurdity of the conven­tional extension program. Admitting that he had been defeated,and that his old service was about to be integrated into nationalproduction programs, he spoke of the day in the hopefully not toodistant future when a new extension service could be created,alongside the old. The latter would continue to work in fomentoprograms. The new staff would pick up again and carryon the" t" I" t " . bessen la ex enSlOn JO •

Three months earlier and 4000 miles away, at the other endof the study area, the incumbent director of the Guatemalan exten­sion service was voicing much the same sentiments as the ex­director in Paraguay. He too was reluctant to see his program

117

lose its identity and purpose in na tional promotional campaigns.This interesting and conscientious man put his finger on one of thecauses of the difficulties that had emasculated the extension pro­gram and left it powerless to accomplish its "essential" job: theabrupt and total withdrawal of American support for one of its ownproducts. He said that he could have survived better if the U. S.had not dropped all moral support along with its technical andbudgetary aid. As it was, he now had to overcome opponents notonly in the ministry but in USAID as well. He thought it was aqueer way for the U. S. to do business, and I think he thought Ithought it was too. If you talk with U. S. agricultural officers inmost USAID missions in Latin America, you will find that collective­ly they represent one of the strongest foci of opposition to conven­tional extension systems. Their judgment, the study finds, iscorrect, but the change in the U. S. position is not enobled by thefact that the U. S. walked away from a problem it created, insteadof correcting it.

The U. S. extension model that American technicians attemptedto transplant in Latin America perhaps could have succeeded underconditions similar to those obtaining in the U. S. in the 1940s and1950s. The land grant structure was firmly rooted in the U. S. bythen, there was a backlog of new technology to disseminate, thesystems of rural credit and farm supplies had extended to allcountries, most farm families had one literate member, andextension agents could count on a wide radio audience. I overheardone of the older, better known, Latin American administrators ofagricultural institutions declare that the transplant would have donebetter if the technicians who brought it had been persuaded by anearlier American school of thought, and worried less about methodsand more about improvising production programs for farmers whohad no credit, no markets, no neighborhood fertilizer dealers, andwho were visited rarely if at all by professional agricultural peopleother than the agent himself.

B. Prescriptions

The institution building model allows for abolishment of in­stitutions that have outlived their usefulness:

"the survival of specific innovations is not necessarilydependent on the continued existence of a given institution.Other institutions may become the receptacles and

118

protectors of the new values and norms. The originalinstitution may have come to the end of its useful socialfunction and its abolishment may become both inevitableand desirable. 1'23

Should the existing extension organizations be abandoned?Not at all~ The network of rural agencies, and the organizationalelan of the services in many areas, is a rich resource that can beharnessed for rural development. What must happen is the com­plete integration of the extension capability in a developmentprocess. This will probably mean the removal of old-school ex­tensionists where they are in a position of leadership and of a mindto resist integration. The study does not point the way toward anoptimal form for the future relationships with credit agencies, re­search stations and other rural services. But extension methodsspecialists have to be demoted in the extension system and personswith broader professional interests and training, that may includeexperiences in extension as well as on these other staffs, allowedto program the activity of the extension agents.

Extension services would be financed according to the re­quirements of the extension function explicit in the production pro­gram, much as the agents included in the vertical wheat programin Bolivia are now budgeted. Single crop functional specialtiesoffer advantages during particular campaigns, but the in-servicetraining program ought to aim for agents who gradually build uponone specialty to a thorough familiarity with a number of crops andproblem areas. Resources will remain scarce, so agents willhave to be concentrated in a few high priority regions and programs,hopefully those with the greatest multiplier effects. The agentswill try for the breakthroughs that are likely to precipitate others.Geographic concentration seems desirable, since, among otherreasons, it would eliminate part of the transport budget. Manyagents could then get along with bikes and scooters. 24 Other ex­tension programs, including much of the work done with ruralyouth and home economics, will have to be starved of funds, and

23Milton J. Esman and Fred C. Bruhns, Institution Buildingin National Development: An Approach to Induced Social Chan e inTransitional Societies Pittsburgh: University of Pittsburgh, Inter­University Research Program in Institution Building, 1965), p. 10.

24The emphasis on jeeps in the early programs may havebeen a mistake: an instance of Americans starting something on animpractically plush scale.

#f,.~

,,,-ilJl

I'

119

the agents' attentions, until such time as the extension organizationcan turn again to non -production goals. Whether the extensionbased programs should be oriented to Indian communities, smallfarmers, or middle size farmers depends on factors the study wasunable to investigate. The author is deeply concerned with thepotential unemployment problem in small holder areas, but under­stands how a dramatic program of high visibility oriented to middlesize farmers might in the end engender more small farm activitythan a program oriented exclusively to the small farmer.

The extension service can play its role as part of an agrarianreform movement, as part of a highland concentration program, aspart of a rice development program, etc. etc. It will lose itsidentity, and many of its so-called clientele, but it will have moreof an impact, attract more attention and may eventually find itselfsufficiently financed and equipped to move back into the desertedareas. Exceptions would be in those countries where the nationaladministration is dedicated to rural progress and willing to fundproduction programs in all areas, such as Mexico. Such excep­tions hopefully will become less rare.

The loss of institutional identity will be hard to accept, but itwas a mistake to begin with and the mistake has to be corrected.

AID's role can be to help, through sector lending, to financethe integrated production programs. Only in extreme situationsshould AID encourage governments to exclude the extension servicefrom such programs. The AID program could usefully includesome measures to improve the salaries or non-salary benefits ofthe agents, but, if this is impossible, should at least help under­write the enormous technical training program that is called for.Commodity imports such as jeeps and cycles are one easy way forAID to cover part of the budget and allow the host government toshift its own budget resources to salaries and other local costs.

The extension function is essential to the rural developmentprocess. The U. S. is responsible for creating an extension capa­bility in the study area. The proper thing now is to program thatresource to perform the essential function. The benefits can makethe sunk costs seem trivial.

120

C. Final Remarks

I am persuaded by this experience to find fault not only withthe conventional extension system, but with the conventional attitudeabout institution building. Just as the extension organization gotout ahead of the information that had to be extended, and the pricingprograms that alone could make it work, so some institution build­ers get out ahead of the programs that have to be institutionalized.Institutional models are constructed which preserve the internalstructure but omit key bridges with other institutions and economicsupport. This has happened elsewhere in the agricultural sector.An example is the cooperative movement, which has created farmerassociations where there is no economic basis for association. Inthe case of extension, the U. S. tried to institutionalize the functionbecause institutions and the function both appeared important. Buta better understanding of the rural development process in theLatin American scenario might have suggested that this was one ofthe functions that shouldn It be institutionalized.

It was said that the enduring benefits of technical assistanceresult from the processes that are left behind. But this study hasfound an instance, and there must be others, when institutionaliza­tion was ill-advised. Perhaps in these instances the programshould be in the hands of the bean breeders, the subject matterspecialists (though admittedly only extraordinary persons ought toreceive such appointments), rather than professional institutionbuilders (unless the two capabilities can be offered by the sameextraordinary pe rson) . The institu tion building specialists, justlike the extension methods specialists, would play very crucialroles on the staff.

The results of the study I believe justify the ambitious studydesign. Broadening the evaluation to include (while distinguishingbetween) effectiveness and Significance criteria, to incorporate theexperience of twelve to sixteen countries, and to begin the studyat the beginning of the program, has allowed us to reach conclu­sions about the overall extension effort that would otherwise nothave been possible. Had the study focused exclusively on thesignificance issue, it might have ended up condemning the entireextension process instead of the peculiar organizational formadopted in the study area. Had the study looked just at El Salvador,or Colombia, the results would have been strikingly different, fordifferent reasons. Had the study started with the situation in 1958,

·1

I,

r

,;."

•

121

which is about the time the most accessible and useful periodicalAID reports were established (the Country Program Book series,the E -1 project descriptions, the End of Tour Reports, and theprinted staffing patterns), the results would have been practicallyworthless. The Agency can indeed learn from its past. The recordis available, though scattered and dusty. Individual AID missionsmay not be the appropriate place to develop such in-depth research,but without the missions help, studies of this sort would beimpossible.

I have been reluctant to alienate any of the readers by over­stating the potential role of the economist in programs such asextension. But one thing is obvious. The absence of an economicbasis for most extension operations, that is, the failure to buildthose operations upon production programs which were both feasibleand profitable, may have been decisive. The fact that professionalextensionists were not linked to professional economic analyticaltalent ranks along with the underdeveloped extension -researchlinkage as one of the principal weaknesses in the extension institu­tion building experience.

If the extension services failed to observe elementary econo­mic laws, the fault was hardly their own. Until recently, practicallyno attention has been given by either AID, the other donor agencies(including the Rockefeller Foundation) or the host governments toapplied agricultural economics - - to the investigation of the profita­bility of modern technologies under varying socio -economic regimes.Unless this gap is closed, no extension operation, however institu­tionalized, can be expected to have a continuing and substantialimpact on farm technology.

-1

• APPENDIX

(A)

MAPS OF THE STUDY DISTRICTS

123

124

18

..10 20 40 Mlle.

I ,

1'0 2'0 4'0 Kilometers

---International boundary

_.- Departamento boundary

@ National capital

o Departamento capital

qo .OUNDAIIY ...EP....NTATION ••NOT N.C".AIIILV AUTHOllnTATIVI

C 0

,-./

r----~.-~.§:::~'-o~.

PACIFIC OCEAN

1. QUEZALTENANGO ,

92

1. PeMn2. Huehuetenanao3. San Marcos4. Totonicapan5. Quezaltenango6. Reta Ihuleu7. Quiche8. 50101019. Suchitepequp.z

10. Chimaltenanao11. Sacltepequu12. Escuintla13. Alta Vera paz14. Baja Vera paz15. Guatemala16. Santa Rosa17. EI Progreso18. Jalap.19. Jutiapa20. Izabal21. Zacapa22. Chiquimula

- 14 --+------"""..d~

Base 59738 9-68

2. JUTIAPA-JALAPA, GUATEMALA 3. NORTHWESTERN EL SALVADOR

"

, -- .C.rtlln 1.I,nd, of the

ArchlpI41••o d. S.n And,••y Provld.ncl. (130 00'N81 0 30'W) .nd tho III. dl~.IPllo (3058'N 81 035'W) I

betonalnl to Colombl, .r.not Ihown on this m.p.

: ':\ (; f I- { C

0('/·,'.1 :Y

125

4. BOYACA, COLOMBIA

.IL NETHERLANDS"' ANTILLES

,~WILLEMSTAD

" ?

GALAP"OOS ISLANDSo 50 100 Mil••

6.50 ~oo Kilomelers

I' . f I,.. ,r

" \' sa~Salvador

F.r~"ndlna'" "': I anla Cr~z

,: 5.0'1Crlsl6ba

l,

u';

ISLALA PLATA

Tala

Base 58056 4·68

126

78

5. AZUAY- CANAR, ECUADOR

76I

aOUNDARY ".PIU:••HTATION ••NOT Nlcu,a"R'I.V AUTHORITATIVI

COLOMBIA

u

ECUADOR--- International boundary

-'- Provincia boundary@ National capitalo Provincia capital

-+-+----+ Railroad---Road

o 50 100 MHesf-I-,.....-"'-,,.----',-,,.---......,

o 50 100 Kilometers

..

127

72

7. PoNO. PERU

Callao is th. copitol 01 the ProvinciaCons/itu<;onol d.1 Col/oo whim hostJ.. slG~s 01 0 o.portom.,,10 but ~too "moJl to b:- sltown on thi, mop.

.aUHDAII..........TATIO" ••MOT Nac",AIIILY AvrMO""'A"~76

6. MANTARO' VAloo mm,

PERUInternational bounda/)/

Departamento bounda/)/

National capital

Departamento capital

Railroad

o

o C E A N

--- Road

0f-I_ ..............,-......:I.l.?O=--.-__.....:;2?O Mil..o 160 260 Kilometers

PACIFIC

0·-----

16

12 -+--------..::n.;;:iBIf':"ir.7'~-~""~~I ....,..""-t_4-:...-

.8 -'F-=:=~

Base 76870 3-70

76091 !Hi9

8. LA PAZ, BOLIVIA

128

BOLIVIAInternaticmal boundary

--- Departamento boundary@ National capital• Departamento capital

--+--+- Railroad

--- Roadc:::J Salt flat

9. UPPER VALLEY, COCHABAMBA, BOLIVIA

APPENDIX B

FORMS

1. Director's Worksheet

EVAL UATION OF THE OFFICIAL EXTENSION SERVICE

(Information to be gathered at Extension Headquarters)

1. - BUDGET

A. Official Budget of the Extension Service (includingagencies, specialists, information)

1960 1961 1962 1963 1964 1965 1966 1967 1968 1969

AuthorizedExpended

B. Extraordinary Sources of Funds

President's OfficeLocal GovernmentPrivate EnterpriseAIDFAO, CARE, etc.IBRD, BIDInt'l 4-HNat'l 4-H CommitteeOther

C. Budget Items: 1968

Authorized Expended

Total

SueldosGastos de administracionAdiestramiento de personal,Equipo, material de trabajo y

transporte129

1968 1969

1950 1955 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969

C. Number of Field and Headquarters Personnel, 1969

1968 1969

130

Number of jeeps at all AgenciesNumber of bikes, etc.Average age of jeepsNumber of Agents without jeepsAverage monthly gas allowance (per agent)

Agents(number of foreign agents)Auxiliaries4-H AssistantsHome EconomicsHome Economist AssistantsPeace Corps VolunteersOther Professional

Field

196719681969

DirectorDeputySupervisors

Headquarters

B. Number of Agencies with a Resident Agent at least4 Months

D. Mate rials

A. Number of Agencies Open

II.- STRUCTURE AND SALARIES

Agencies:Sub-Agencies:Regional Centers:Other (SFEI, etc.):

131

ChiefRegional4-HHome EconomicsOther (name)

Specialists (name)

D. Educational Level of Agents: 1969

Ingeniero AgronomoAgronomoPerito

E . Salaries :Starting

Direct from School Second Yr. Fifth Yr.

Extension Serv. Basic

AgentIngenieroAgronomo

Research Service

IngenieroAgronomo

Agricultural Bank

IngenieroAgronomo

Private Sector

IngenieroAgronomo

F. Extra -Official Jobs: Agents

Permitted:Condoned:(% of Agents)

Extra Basic Basic

···-----------------EL

132

G. Turnover of Agents

Total Agents Ingreso Abandono

19651966196719681969

H. Number of agents who enjoyed political appointment topresent position.

III. - TRAINING

A. Average number of weeks (per year) per Agent

OrientationIn-Service

B. Average Number of professional staff in OverseasTraining (man-years)

IV. - CONTINUITY OF DIRECTORSHIPExtension

President Minister Director Deputy

1940

1945

1950

1955

1960

1965

1969

133

V. - PROGRAM CONCENTRATION

A. Geographic:

B. Crops

Departments Municipios

Number of Priority CropsExtension ServiceEach Agent

VI. - PRIORITIES

A. Distribution of additional funds (assume 25% budgetincrease)

Increase salariesOpen new agenciesIncrease staff in present agenciesIncrease specialist staffIncrease supervisory staffVehiclesOperating expensesOther materials

educational. vis ual aids. etc.seeds. fertilizer. etc.

VII. - COORDINATION

A. Annual Conferences

1967

National Level (yes /no)Regional Level (yes /no)% Program devoted to

technical information%of that time conducted by

research personnel

B. Credit Programs

1968 1969

Extension Service Banks

Small farmers covered (%)Small farmers Delivered to Banks (%)

iI

.~

I

~

1

~~(,

•,1!!

4i

'¥

1'I

~,i

~

Ii

"~

j

~,1j

i~

",f't

~

J

APPENDlX B-2

PROFESSIONAL QUESTIONNAIRE

Country:

Neme of Interviewee:

Position:

QUESTIONNAIRE

EVALUAtION OF THE OFFICIAL EXTENSION SERVICE

This is part of a study of the Agricultural Extension Services in 16countries in Central and South America. It deals with certain insti­tutional issues for which personal judgements are more important thanquantitative information. The same questionnaire will be used in eachcountry: thus a few of the que stions may not apply in any particularcountry. It will be given to ten persons in each country. usuallypeople working outside of the government who have an interest in theproper functioning of agricultural extension and a familiarity withlocal conditions.

Your answers will be treated ·as confidential, and we hope they willreflect as frank a judgement as you can make on these qualitative in­dicators of institutional viability.

For each question (except number 9) please give a rating from 1 to Saccording to the following scale. or check the column indicating yourinformation is insufficient to permit a fair evaluation.,

~ Outstanding4 Good3 Fair, just adequate2 Poor. inadequate1 Very poor, none

E. B. aICIAID/PrC/WashingtonSeptember. 1969

135

a. On large commercial farmsand ranches:

(1) ~nich are visited by the -----Agents.

(11) Which are not visited bythe Agents. ----- -

b. On medium-size farms:

(i) Which cooperate with theAgents (the diress effect) -----

(i1) Which do not cooperate, butare in the same general area(the indirect effect) ....... -----

c. On small-size farms:

(i) Which cooperate with theAgents.

--~--

(ii) Which do not cooperate, butare in the same general area. -----

IMPACT Don'tKnow

Rating1 2 3 4 5

Please indicate t.he tmportance of theextension service in tmproving farmtechnology:

136

1. In communities given preferenceby the Agents, to what extent 40the farmers appreciate the Ex ­tension Service?

2. In communities given preferenceby the Agents, to What extent dothe farmers trust the technicaladvice provided by the extensionservice?

3.

Rating

137

IJ:ADERSHIP AND PRESTIGE

4. How high is the loyalty of extensionagents to the extension service adminis­tration?

1 2 3 4 5Don'tKnow

s.

6.

How much agreement on the objectives andoperations of the extension service isthere between the supervisors and thedirectors?

How high has been the prestige of theextension agents within the professionalcommunity during the last few years.

7. How high has been the prestige of theextension service's directorShip withinthe govermnent during the last few years..... _

8. How high has been the prestige of theextension service's directorship withinthe private professional community.

9. If support 1s needed by another agencyfor a project proposal that required thehelp of extension agents, to whom shouldthat agency appeal for quick approval andaction (underline two): President, Minister,Sub-secretary, Director General, ExtensionDirector, other (please identify the other

)

RELATION WITH OTHER ORGANIZAIIONS

10. How would you rate the extension service'scoordination with agricultural researchinstitutions during the last few years:

a.

b.

At the Director's level

At the Agent's level

1 2 3 4 5

a. At the Director's level

Don'tKnow

At the Director's level

At the Agent's level

a.

b.

The degree to which those objectivesagree with national agricultural goalsas expressed by the National Plan (orby the Minister, if there is no plan).

The degree to which the extensionservice has had a set of clearlystated objectives during the lastfew years.

The degree to which the extensionservice has tried to accomplishthese objectives.

138

b. At the Agent's level

Rating

PROGRAM

15.

12. To what extent doe s the extension servicecoordinate with other "fomento" institu­tions which deal with agriculture?

14.

16.

11. How would you rate the extensionservice's coordination with agriculturalbanks which provide credit to the typeof farmer with whom the extensionservice works?

13. To what extent do other institutionsnow request the help of the extensionservice in carrying out their ownprograms

17.

18.

19.

139

Rating1 2 3 4 5

The degree to which the extensionservice has demonstrated a capacityto develop new programs during the lastfew years without the need of foreigntechnical advisory assistance.

The degree to which it has demonstrateda capacity to promote the funding ofnew programs among non-traditionalsources of finance.

The degree to which these new programs,or improvements in existing programs,were the product of a deliberate ap ­praisal by the extension service of itsown experience (the "feedback" mechanismL _

Don'tKnow

APPENDIX C

RELATED ISSUES IN INSTITUTION BUILDING

1. Regional Demonstration Programs and Local Governments.

The study included within its scope six regional programs inwhich the extension service had an important role: STACA Project#3 in Boyaca, Colombia; Plan LAC in Azuay-Canar, Ecuador; theMantaro Valley Development Plan in Peru; the Southern Peru De­velopment Program; Santa Cruz, Bolivia; and Plan Chillan, Chile.There are others, for example the Cauca Valley Program inColombia, but these six constitute the majority of the big ones(outside the settlement zones). All six originated as servicioprojects.

From readings and interviews with veterans of these pro­grams, I have reached a general conclusion as to the effectivenessof each. I cannot substantiate these impressions, and other ob­servers may disagree with one or more. My advantage over mostother persons is that I have studied them all, and been able to makesome comparative judgments.

Of the six, I consider that Southern Peru and Boyaca were un­successful, Plan LAC, Mantaro, and the extension phase of PlanC hillan were qualified successes (the research phase of Plan Chillanreportedly was excellent), and Santa Cruz was a beauty. Santa Cruzwas unusual, however, in that it enjoyed an enormous infusion ofpublic and private capital.

The Boyaca story is incredible, and illustrates a problem thatfouled up at least three of the others. The Boyaca area was selectedby STACA for a pilot extension project partly because the area wasso poor, and partly because the then President of Colombia --Rojas Pinilla -- was born there. This was about as pure an exten­sion project as obtained in the study area. There were some subjectspecialists, including four Americans (horticulture, livestock,agronomy, forestry), working in the area with the agents during theprojects' six-year life span. But the thrust was toward the crea­tion of an extension organization, the American extension advisorwas project director during the first four, crucial years and theemphasis of the training program for Colombian agents was on ex­tension methods, not technology. The project was intended to (1)demonstrate the importance of extension; (2) train Colombians for

141

142

work in the future national extension service and (3) increase in­come s in the area.

The fact that the project had very little impact on incomelevels is discussed elsewhere in the general report. The trainingaspect was successfully concluded, at least to the extent of pre­paring a cadre of Colombians skilled in extension methods. Thedemonstration effects were ludicrous.

Although a national extension service was created in 1958 outof the Boyaca beginnings, the fact that it was in financial difficultiesalmost from the start, and later died (the only U. S. built extensionservice not to survive), lends support to the judgment I was givenby several Colombians that this was never really transformed froman American initiative to an indigenous program. As early as1963, there was serious discussion of moving extension out of theMinistry, where the Americans helped put it. When ICA took overextension in 1967 it dropped all but six of the inherited agencies.

The worst of it, however, was that the government in the stateof Boyaca did not want the project to continue in the state afterphase-out of STACA activities, or at least gave it so low prioritythat the state funds were insufficient to sustain it. An agreementbetween the new national service and the Boyaca state governmentin 1960 left extension in that state entirely in the hands of the stategovernment. Two years later the state government closed theagencies and dispersed the vehicles, typewriters and sewingmachines to other units. Thus, from 1962 until 1968, when rCAagents moved back into Boyaca, there were extension agencieseverywhere in Colombia except in the area where extension began.r interviewed the American extension advisor who headed up theBoyaca project from 1954 to 1958, and found that he was uninformedof its fate.

The lesson is that regional development programs cannot sur­vive unless they are acceptable to local governments, and localgovernments must be vigorously wooed. Plan LAC1 got hung up onthe same problem. Norman Ward (U. S. Agriculture chief andservicio director), and the Ecuadorian administrators of Plan LAC(who were not local people), got into a feud with the Senator from

1Plan Azuay-Canar ran from June 1956 to September 1957.It was followed by Plan Austral, which name was changed to PlanLAC in 1958 to reflect the inclusion of Loja in the perimeter. Irefer to this entire sequence of operations as Plan LAC.

,l

,-iT

143

Cuenca (the capital of the area and headquarters of Plan LAC) andother local officials and were eventually forced to close the projectprematurely. A local organization, CREA, replaced it, absorbingsome of its programs and administrative practices. CREA hassince earned an excellent reputation for its fomento type programs.But it is hardly an outgrowth of Plan LAC, since the material re­sources and manpower of Plan LAC were hastily withdrawn (partlyto avoid their falling into the hands of the rival organization). PlanLAC began some important agricultural programs which continueto this day. The poultry project, described in the general reportstudy, is an example. Plan LAC failed to raise productivity levelsdramatically, but as a demonstration project it must be ratedrather high. Still, one can imagine how much more impact PlanLAC could have had if the SCIA personnel who directed it had en­listed the local administrative talents, which evidently were pres­ent, rather than fight them. I recognize this is easy to say frommy vantage point, and may not have seemed so easy in Cuenca in1956 when SeIA first moved its team in.

One of the reasons the Southern Peru program was abortedwas because SIPA officials could not get along with the new autho­rities in Puno, especially the leadership at the new university inthat city who had inherited the local agricultural experimentalstations. SIPA was the government entity which was formed fromSCIPA, the servicio, when the latter was abolished in 1961. NorthCarolina contract technicians posted to Puno were able to developworking relations with the university staff. But without supportfrom SIPA, one of their host organizations, these technicians wereunable to effectively "extend" from the university, except on theirown and piecemeal. This is the same sort of problem.

Plan Chillan created opposition it could have avoided when itelected to set up a new extension operation in the three provincesin Southern Chile where the Plan was operative, rather than usethe institutional structure already available. I have talked withthe then director of the national extension service based in Santiago,a man who held that position from 1954 to 1960 (Plan Chillan ranfrom 1953 to 1960). He lost some good men to the Plan. Theywere not stolen from his service, but he had to give up effectivecontrol of their time. Funds that he claims he could have usedbetter in other areas were diverted to Plan Chillan against hisadvice.

144

The Plan, he admits, had practical value in changing tech­nical practices in the region. Other commentators split about halfand half over whether to call it a success. The agricultural facultyat the University of Concepcion undoubtedly benefitted a great deal -­it was started under the Plan and inherited the Chillan station.The Chilean legislature liked it, and continued it two years afterU. S. phase-out. One of Plan Chillan's selling points that dis­tinguishes itself from servicio operations in other countries wasthe deliberate effort to use Americans only as advisors, reservingthe operational positions for nationals. But Plan Chillan was con­ceived as a demonstration project, and since the leader of thenational extension service should have been one of its principalwitnesses, his estrangement from the Plan was an avoidable error.

I was unable to draw any conclusions about the efficiency ofregional demonstration programs per see In principal they makesense. However, local governments have to be involved, and thefailure of the servicio to adequately cover this flank probably ex-

•plains why the U. S. expe rience in the study area is not particularlyimpressive.

2. Servicios as an Instrument for Institution Building

Figure 3 on page 33' shows the life span of the agriculturalservicios in the study area. Legitimate extension services originatedwithin the servicios in all cases except Chile, which had an exten­sion service long before the U. S. technicians arrived, and EISalvador and Panama, where Americans helped create the servicesbut these remained outside the servicio structures.

I say Illegitimate, " because in six of the other countries afarmer service already existed in whatever ministry handledagriculture, and this farmer service was usually referred to' asextension. The servicio extensionists disqualified the older organi­zations becaus e the latter were oriented toward fomento. It is aa fact that they did engage in regulatory and promotional activitiesand participated in the distribution of supplies. But it must havedisturbed many of the ministry officials of these older office S tohave their job condemned, so to speak, by the Americans and tolose their credentials.

Of the six "illegitimate" services, five had been started byAmericans in the pre servicio period of aid. I talked with aU. S.

145

extensionist who was in Colombia during 1947 and 1948 under aUSDA program. He helped build a national extension service there.It was in the Ministry of Agriculture and labeled the Division deExtension. And indeed this division continued to operate until the"national extension service" was formed by STACA in 1958. Hewould be surprised to learn that the service he helped create wasdiscredited in 1953 when STACA began Project #3 in Boyaca, andthat fellow Americans ten years after his departure dated exten­sion in Colombia from the day they arrived rather than from the dayhe arrived. Fortunately he had never bothered to find out, and wasunaware of his "failures. "

I make this point because it sets a stage for the issue I want toraise concerning servicios. The use of joint operations, likeservicios, in aid affairs is controversial. I have had to study them,because their history is inextricably linked to the rise and fall ofthe extension services. In general I feel that they proved to be anexcellent device for getting something done, and that it was possibleto phase -out the U. S. contribution without destroying or weakeningthe action programs and the institutional structures we wanted thehost government to inherit. In Peru and Honduras this in facthappened: the programs and structures were preserved in autono­mous agencies run by the local government. These agencies havejust lost their autonomy (SIPA was dissolved in 1969; DESARRURALin 1970), but the programs and vital parts of the structures survive.Moreover, the Rockefeller Foundation has been successful in leav­ing behind thriving, semi-autonomous research organizations inseveral countries. Thus phase-out of the foreign partner from ajoint venture need not upset the institution-building process (thoughin some situations, e. g., in Bolivia, where the SAl staff was triplethe size of the ministry staff, it is hard to imagine how the transfercould have been successfully managed without some difficulty).

The experience with extension services, however, suggeststhat the people who run servicios have usually been unable to pre­pare for, and effect, successful transfer. Linkages with the olderministerial divisions, with which extension services had to coexistafter transfer, were not developed. The linkages between con­stituent units of the servicio -- between extension and research forexample -- were themselves not sufficiently strong, and in mostcases these units flew apart once they left the protective custody ofthe servicio. We can benefit from the IB analysis, and the empha­sis it gives to linkages. A servicio project scheduled for transferwants to make friends and avoid enemies in the entrenched bureauc­racy. IB indicates that the mistake may have been in designing

146

organizations for work in the servicio. AID should have been de­signing for organizations that would survive in the ministry.

Opposition to servicios arose in Washington from persons whowere concerned about this very problem. They argued that servicioprojects rated well on performance criteria, but failed to meetelementary institution building tests that were obvious long beforeIB was available. They demanded the rapid close-out of serviciooperations, and had their way. I don't know whether the blame forthe subsequent mess belongs with the servicio directors, who sup­posedly failed to build for the future, or the AID/Washingtonofficers who saw part of the problem and called for prematurephase-out.

But the issue is bigger than that. In a short paper he wroteabout servicios in 1958, Edwin Astle listed the questions a missionstaff should ask before accepting proposals for servicio projects. 2These were:

1. "Learn enough about local institutions, whethergovernmental, private, cooperative, commercial,administrative, or otherwise, to know where thegaps are in the country's total operational structure.

2. Having developed your program and divided it intoprojects and established priorities on project execu­tion, take a look at the local institutions and deter­mine which ones have interests that relate themselvesclosely to the projects which comprise your pro­gram, to see if anyone, or any combination of them,could logically carryon the activity you have in mind,either now or in the future.

3. If the answer to that is NO, then take another look atthe activity to see if it can be carried out in a dif­ferent manner, or arranged differently, stillachieving its objectives, but fitting, now or eventu­ally, within the possible scope of an existinginstitution.

2Edwin P. Astle, HAgricultural Servicio Programs and Timingand Technique of Transferring Agricultural Servicio Programs toHost Countries" (lCA, January 7, 1958), pp. 4-6.

147

4. If the answer to that is still NO, size the projectup again to see if it would be worthwhile for a newinstitution to be established or an old one reorganized,to eventually carryon the activity.

5. If the answer to that is still NO, take another look atthe project. Take a long look and a good one, becausethat's a project you'll have for a long, long time. Ithad better be good, if you keep it beyond this point.

6. If the answer to any of the above questions is YES,then is when you determine whether you need aServicio at all, and whether the project you have inmind should be born within, or outside, the Servicio.To make that determination ask yourself the follow­ing questions:

(a) Do the responsible people who program for orwork within the institution agree with the objec­tives and want them to be carried out?

(b) What keeps the institution from wanting to, orbeing able to, start carrying out the objectivesnow? If it is simply the need for technicaladvice, shouldn't it be handled without becominga part of the Servicio operation?

(c) If it is determined that the gap can't be filledsimply by providing a technical adviser, whatchanges need to be made in the institution toget it in shape eventually to assume operatingresponsibility for the activity?

I/1

(d)

(e)

(f)

Does the institution want to make those changes,and can they be made? How long is it likelyto take and what steps are involved?

Is the project so designed as to facilitate themaking of the changes in the institution and fillingthe gap which the project is designed to filltemporarily?

Is the project so designed that it will be possibleeasily to judge when the institution can assumeoperating responsibility for it?

148

(g) Can you make a success of running the activitywhile the institutional gap is being filled?"

Astle's principal point is that servicios should never take ona job which local institutions could handle, or could handle inanother way with the proper support. The U. S. organized serviciossimply ignored that kind of advice. The fact that fomento orientedextension institutions within the governments were dropped in favorof new extension programs in the servicios violates this basic con­cept. The fact that the "program space" was already occupied didnot bother the U. S. architects, and it should have. It is all themore distressing because the orientation of the new programs wasless productive than the old. and the U. S. probably would have donebetter in the productivity sense as well as in the institution-buildingsense by improving upon the existing institutions than scrappingthem.

If the level of competence in AID missions is such that it can­not better identify the projects suitable for servicios. and bettereffect transfer of them later on. then perhaps AID should stay awayfrom servicios despite their recognized advantages. This is not in­tended to run down the "level of competence in AID missions." Iam convinced that identification and transfer require the highestcalibre of social engineers, and these men are rare. The fact thatit has been done well on several occasions cannot all be attributedto luck. It means that successful servicio operations are withinreach, though they cannot be assumed.

A good tip was offered me by Samuel C. Litzenberger. anagronomist in Nicaragua during the STAN period and. according tomy interview notes, the most respected agricultural scientist theU. S. ever sent to that country. He said that servicios are veryuseful for performing demonstrations, and for getting things goingfor which the government has little initial interest. The trick is,after the pilot operation. to convince the government that it was thegovernment's idea and program all alone. That's quite a trick.Litzenberger knows. STAN was suddenly abolished in 1958, follow­ing the accession of a hostile politician to the ministry, and part ofthe U. S. team was thrown out of the country.

I must note here that the attitude of the host officials withwhom I spoke about the servicio issue was also mixed. Thirtyofficials. ten of whom had been outside the servicio, looking in.were asked whether they thought the joint operations had been. on

.iI

,

f

i"

'~

149

balance, a good thing. Fifteen said yes, eleven said no, and fourcouldn't decide. For the ten outsiders, the figures were two, sixand two, .~howing, as we might have suspected, that those who werein the servicio liked it and those that were not, didn't. A few of theopponents were convinced that it would not have been as difficult asthe Americans claimed it was to create an effective extensioncapacity within the host government. One cited the Panamanianexample, where the OFAR sponsored University of Arkansas teamhelped develop the service within the ministry (and outside of the(rival) servicio). An excellent Panamanian was appointed to headup the new service, and he was effective, despite "red tape, " untilhis early, untimely death.

3. U. S. Extension Advisors

This report is critical of the extension institution buildingprocess. It is not critical of the U. S. advisers who were broughtto the study area to set it in motion. Of the 576 individuals identi­fied on the twelve charts similar to figure 1 the author has inter­viewed 143 (there is some double counting). Of these, twenty-eightwere extension advisers. Most were personable, well informedtechnicians. They knew the peasants of Latin America. Not enoughof them spoke good Spanish, which is unfortunate, since extensiondepends on communication and the adviser ought to be able to com­municate when he is teaching the sUbject. In fact it is a mystery tome how persons with no facility nor interest in learning Spanishwere sent on these programs. Many did speak Spanish well, how­ever, and a number of the key advisers were Americans of Mexicanextraction.

During interviews with twenty-three host country extensiondirectors (incumbents and predecessors) and other officials, Iasked for their summary assessment of each of the U. S. technicianswith whom they had been associated. They were asked to rate thesepersons either outstanding, good, fair and poor, and to give equalweight to technical competence on the one hand and the ability toshare it on the other. Altogether 112 American extension advisers,subject matter specialists and principal agricultural officers from

150

eight country Missions were evaluated. The results were:

This is a commendable record, because it suggests that only thir­teen percent of the total were unqualified and shouldn It have beensent. Extraordinary technical or administrative competence wasthe factor that usually explained outstanding ratings. Personalitydisorders were usually behind the poor ratings.

I had expected that extension advisers, extension trainingadvisers, home economists and rural youth experts would not rateas well as the technicians who filled subject matter specialist slotson the U. S. extension team. This would reflect the assertion thatextension methods are not very difficult to transfer. Counterpartspick up the principles of extension methods quickly. Once thishappens, the continuing presence of any U. S. adviser becomesanachronistic and occasionally irritating. The results did not con­form to the expectation, however. Of the 112 technicians, 61 wereextension types. Their scores were almost identical with the onesjust tabled.

The problem with the U. S. extension presence in the studyarea was not so much the U. S. technicians involved as that they

11 d "A' h t' d' twere over-programme. mer1cans w 0 were ra1ne 1n ex en-sion methods, and were used to operating out of a county agency ina developed economy, were asked to take the leadership in design­ing and implementing rural development programs for underde­veloped economies. The Boyaca extension program is an example.The approach can work wherever the lack of a farmer informationservice is the only major constraint on growth. It does not work inthe study area. A few individuals, such as the famous John R.Neale who directed SCIPA in Peru from 1944 to 1958 (he was madeMission Director in 1955), broke out of the extension mold andforged an integrated rural development program. But most exten­sion advisers were out of their element. It wasn't fair to give themthe assignment, and the numerous failures described in this study

..~

151

are not their fault. Within the area where advisory assistance inextension methods and structure could be expected to have in­fluence - - in the development of a field organization - - theseadvisers often did admirable jobs. People like Nick de Baca andErnie Gutierrez were pioneers in this field. and extension organiza­tions in several countries have a substantial debt to them. If theorganizations were not used well. the fault lies at another level ofmanagement.

CRITIQUES

The rollowing statements represent the views or their authors andare not necessarily representative or the views or their parentinstitution.

'.'

J. L. MatthewsAssistant Administrator

Extension serviceU. S. Department of Agriculture

The Latin American extension services, like schools and govern­ments, can be and are criticized for their inadequacies. Many Of thecriticisms are fair and others stem from frustrated and unrealisticexpectations. To some degree, this is to be expected from concernedoutsiders who are impatient to see the developing countries achievetheir potential social and economic advancements.

As Mr. Rice brings out, there is much that can be said in favorof the existing extension services. They are functio~ing and

. accepted by large numbers of people in each country. The relativecost of developing these services to their present stage has been lowcompared with expenditures for other less-enduring and less­successful assistance efforts.

In most countries, no other indigenous institution offers ameans of reaching the rural massee with agricultural technology andother instruction that will help them make their contribution tonational social and economic development. Thus, extension servicesare an available instrument of change that cap. be employed fornational development when the governments are ready to establishpriorities and give sustained support to them. They can be theinstruments for relatively quick and high returns from resourcesallocated to rural development.

A comprehensive and thorough evaluation of extension servicesin relation to U. S. foreign assistance efforts to establish them islong overdue. Valid reconstruction to assess the value of U. S.inputs may be impoesible at this stage, but a benchmark would helpto determine the present and future needs.

A considerable number of limited studies of extension serviceshas been made in Latin America during the last 25 years. None hasattempted a study on the scale of this one.

The comments and reactions that are included in the critiquerelate to some of the more important generalizations expressed. Itis not practical to comment on any other than those included here.However, failure to comment should not necessarily be construed asendorsement or agreement.

This study is disappointing in the quality of data and thefailure to obtain enough information about certain factors that areimportant influences on the success or failure of the extensionservice. It is further limited by the difficulty of reconstructing

- 2 -

from skimpy records, faulty and often biased memories. The study washampered by the limited scope of individual respondents' knowledgeand the absence of objective data about what was attempted and theresults that were actually achieved.

Mr. Rice has brought toeether a great deal of interesting in­formation about the aericultural extension services in Latin Americaand U. S. assistance to them. The whole story could not be told inthe time available and for purposes of this study. Generally, thedata are limited for the purposes of assessing cause and effectrelationships in a very complex set of forces and circumstances.

For example, some of the important factors that Mr. Rice hastouched upon include general purposes, financing and governmentsupport, organization and staffing, staff competencies and programsupport, technology and research, education and training of staff,image--public and private, and program impact.

Some others were omitted or little data was provided, such as,the role of extension services in agricultural development, foreignassistance agency and national government relationships, administra­tion and coordination, program development, involvement and leader­ship, reporting and evaluation, and cultural influences.

We must assume some lack of objectivity among the respondentsinterviewed, and in some instances there seems to be considerablebias. Lack of objectivity, always a problem in this kind of study,could bias the information in a variety of ways. The responses areas likely to favor the extension services as they are to be againstthem.

There is reason to question his conclusions regarding how wellcertain Americans performed their roles in support of the extensionservices. This comment assumes that an American assisting an exten­sion service in another country should involve his foreign counter­part in a way that will enable him to function much better after hisdeparture than he would have if the American had not worked with him.Admittedly, developing the abilities of a counterpart is difficult.An American has more leverage than a :(oreign national with the otherA~ericans in the country who control or influence the availabilityof physical and other inputs that are essential for the success ofthe extension service programs. Often the same kind of leverage canbe exercised with the nationals in positions of authority. Theleverage is not a mantle that the American can pass on to the counter­part when the American leaves. The only way to convey it is for theAmerican to teach him to build the relationships that will give himas much as possible of what he needs. In defense of the American,it should be said that often his tenure is too sh~rt to accomplishthis purpose, or his services maybe abruptly withdrawn for reasonsunrelated to the extension service's needs. Furthermore, his personalrewards may not be for what the foreign national does with Americanhelp.

- 3 -

It is questionable that there has been, durine the period ofassistance in forming extension services, goal consensus even in U. S.official circles and amone the American officials in charge of develop­ment in the countries. There is even less likelihood that it existedamong the foreign government officials. Lack of agreement on goalsand policies regarding institutional development and agriculturalproduction goals make drawing valid conclusions about them ahazardous exercise.

Chapter 3 suggests that some extension services appear to havedrifted into activities not closely related to development goals.It is difficult to focus educational programs and action toward endsthat are not clearly articulated at the higher bureaucratic levelsin a country. Even if there is change, hOW can it be related to any­bOdy's goals except the goals of the local extension workers who areinvolved? Meanwhile, the farmers and the extension workers perceivesuccess in terms of what they were trying to do, whether or not itis consistent with goals of the hif.ftcr echelons in the eA~ension

service or the national government.

The main criterion of success employed by ~~. Rice is increasedagricultural production. It is not clear whether he had in mindproduction for domestic consumption or production to earn foreignexchange. The latter is a consuming purpose in some countries. Amongthe poorer small farmers in a country, a considerable increment ofadded production will be consumed by the fa~lies that produce it,thus having virtually no effect on the co~modity markets.

The study contains little data bearing directly on agriculturalprOduction, which is the assumed priority goal for the extensionservices. Unavailable are data that will allow comparisons overtime or between countries. This is especially true of crops producedby small farmers for local consumption.

It should be realized that there may not have been a practicalministry or agency strategy to move the extension service toward theinstitutionalizing goals. Examples are the differences Mr. Ricepoints out among the Brazilian state extension services where thisphenomenon was reported by another observer (1). There seems to bea relationship between the effectiveness in institutionalizing andprogram goals when these are perceived and understood consistentlyboth in the field and at the administrative levels.

Analysts should be cautious in attributing failure to increaseproduction to failure of the extension institution. There may beconflict between the national production goals and the f,oa1s offarmers. At the farmer level, the lack of consensus on goals is mostcritical. The farmer adjusts his evaluation of production potentialrealistically to the economic and social situation in which he mustproduce. Therefore, the national production goals need to be re­flected in policies and programs of all of the government departmentsand agencies that reach the farmer. In addition to a profitable

- 4 -

market, the essentia+ inputs must be available to him if he is toincrease and maintain a higher level of production.

Mr. Rice stated a number of broad generalities about LatinAmerican extension services. He has some positive conclusions aboutthe state of the extension services, such as:

- The concept of extension of agricultural technology isgenerally accepted in the countries.

- They have made contributions to higher status of women,youth, the home, and the community.

- Good will for the extension services has been generatedamong the people.

- The cost to the United states and the national governmentshas been modest.

These factors are important, both in rural development of a countryand in the development of an extension service. Also, they giveevidence of success of the U. S. efforts.

Other conclusions are about the objectives or expectations ofthe agency and the staff involved in foreign assistance, such as:

- U. S. assistance was in the wrong places and uncoordinated.- Assistance was given in ambiguous ways, contributing to the

present state of things.- The extension services are understaffed in both numbers and

quality of staff.- Extension services are not capable of developing their own

programs, which should be prepared at higher levels.- Conventional extension work overemphasizes methodology.

Presumably, the reference to proliferation of extension capabi­lities in other rural development agencies refers to other agri­cultural agencies undertaking extension activities along with theirregular field activities. This could be expected as a normal out­growth when there is no comnnmication, or where there is activecompetition between the agencies at the national level. It candevelop when the extension service administration is unwilling tocooperate by developing appropriate educational programs in the areasof other agencies' interests. For example, the ministry agencyresponsible for farmer cooperatives or agricultural credit, in theabsence of supporting extension programs, may undertake extensionwork on agricultural credit and cooperatives. Coordination andcooperation are two-way streets that should be promoted in theinterest of efficient use of resources and monitored by top-levelministry staff.

Considering the limited resources and competition among nationalpriorities, no developing country can afford to staff its totalpersonnel needs at this time. The money, trained personnel and otherinputs are simply not available for an extension service reaching into

..

- 5 -

It is unrealistic to expect national extension services to cometo full maturity in less than about a generation, even with adequateresources. We know that 25 years ago it took 8 to 10 years for matureextension services in the U. S. Corn Belt to get mOst corn producersto adopt hybrids, one of the first high-yielding crops comparable tothe new varieties of rice, wheat and corn now being spread aroundthe world. It takes time, consistent U. S. assistance, and nationalsupport to build an effective national extension service.

The most visible increases in production tend to be on highlyspecialized large-scale crops that are only possible for producerswho control ample capital and land. Examples are bananas, wheat,rice, cotton, and cattle. Corn, since it is a staple in the diet ofLatins, is produced and consumed in large quantities by small farmersas are vegetable crops. From the market point of view, consumptionneeds compete with the family desire to earn cash for other purchases.Data on quantities of food crops consumed are hard to get and ofdoubtful accuracy when they are obtained, SO that a good measure ofchange in this area may not be possible.

every nook and cranny of the country. A more realistic approach maybe to organize and staff, with adequate support materials and facili­ties, whatever size geographical area is feasible at the time.Furthermore, the area or areas should be chosen according to thepotential for agricultural production or in relation to some otherpriority. They should not be established as "demonstration," "pilot,"or 'any other designation that implies an experimental or transitorysituation. They should be initiated and maintained as a permanentprogram that may expand in time to other areas. This concept isexplained by Mosher (2) in Appendix A of his book on creating aprogressive rural structure.

The question of whether or not the extension service should workprimarily with the large producers or the small farmers has been acontroversial one. Few extension workers abroad could be expectedto have the qualifications required for working with large-scalelivestock or crop operations and the producers who have them. Thus,due to inadequate technical preparation and subject-matter support,most extension workers are excluded from contributing to the potentialsources of large increases in agricultural production. We can assumethat in the Latin American countries a very small percentage of the

Few, if any, developing countries can adequately support oneeffective field organization country-wide let alone several, whichsome are attempting. A single extension service with adequateresources can perform extension functions for crops, livestock andfood production, health education, home improvement, family planning,and youth programs, agriCUltural credit, cooperatives, etc. It canif it has the needed technical support, staff training, and coopera­tion of the appropriate national ministries and agencies. This maycall for a new approach to extension service program planning andinter-aeency coordination.

- 6 -

farms produce the greater part of what goes into the domestic andforeign market channels.

Every country needs some kind of field organization to carrynew lmowledge about farming to the farmers, large and small, and helpthem adapt it to their own use. MOsher (2) has stated these pointsvery clearly in his book on rural structures for agriculture.

In order to accomplish these major purposes, there must be asource of adaptable technology that can be applied in the local situ­ation. The extension service and the lmowledge source must have thekind of relationship that allows a two-way flow. Research informationmust flow through the extension service and be adapted by it to thefarmers' needs. Information on farmers I problems must flow backthrough the extension service to the research organization SO thatthey can be investigated. Too often there is little or no flow ofresearch to the extension service or a one-way flow SO that theresearch staff may be out of touch with farmers' practical problems.These functions can be organized and related in more than one way.

For organization effectiveness, another dimension is significant.Appropriate purposes and functions must be built into the extensionorganization at each level to give direction in action progra~s.

These are necessary for accurate evaluation of operations and results.To say it another way, the extension service institutional conceptand implementation strategy should be clarified among U. S. assist­ance personnel in Washington, in the mission, and in the country fromthe national leadership to the level of extension workers directlyin contact with farmers.

A third requirement is to provide basic technical education andin-service training to prepare the extension worker for effectivenesswith an identified category of farmers. Specific training is neededto reach production and other goals. This is not a kind of trainingthat can be generalized leaving the problems of adapting the researchand the extension methods entirely to the worker at the farmer level.In-service training and backstopping by extension subject-matterspecialists are essential for a highly successful program.

Mr. Rice questions the value of an extension service that is aseparate autonomous institution. This is a good question asked inthe context of arrangements for providing for staff academic educa­tion. It can be asked about a source of research on crops andanimals that can be adapted to the various farming areas in thecountry. It is an appropriate question in terms of sources of creditfor farmers and the availability of the physical inputs that areneeded. No single organization will suffice for all of theseresources and the essential physical inputs. Failure of functionsresult from many causes not inherent in the organization or institu­tional arrangements. He has described some of them.

..

- 7 -

I consider erroneous Mr. Rice's conclusion that extension pro­grams must be written by authorities higher than extension. I am notaware of a single successful instance where the extension serviceprogram was planned at the national level onl~r. Rather, prioritiesand the goals should be established at the national level, thenthese should become a part of the basis of extension programs at alllevels in the service. In this way the readiness of rural peopleand availability O.f resources in any particular area can be takeninto account in planning and implementing the programs.

The lack of adaptable technology has plagued the extensionservices in Latin America from the beginning to this time. The pro­blem may be little or no research, the wrong kind of research, or .lack of a two-way flow of research results to the ~ctension serviceand to the farmer, as well as the absence of the flow of farmers'problems in the reverse direction. Aside from this, there is thelack of credit, price incentives, market systems and physical inputs.

Verification trials by extension aeents or subject-matter speci­alists are often overlooked as an extension service capability thatcontribute to adapting new crops and practices. If the agents arenot capable of conducting these, then the subject-matter specialistsshould supervise or actually conduct them.

Although the study includes much pro and con, it has not shownthe "Extension concept" to be faulty or inapplicable in foreign situ­ations. What it does show is the futility of establishing an exten­sion service without assuring the resources and other needed supportlong enough to institutionalize it and its function. This meansproviding in addition to money, available research and physicalinputs.

Considerable space is devoted to discussion of "attempts totransplant" the U. S. Ex:tension Service model. There are some basicideas or principles that have led to success of extension servicesin this country. None of these require a particular organization orinstitutional arrangement. Some relate to government or institu­tional policies. Others have to do with functions or operationalarrangements. Some require a period for development or maturity inorder to achieve them. Some of the more relevant and important forthis study are discussed briefly below.

U. S. Extension Services have maintained a non-political postureand avoided regulatory, legislative, and administrative roles withnational or local governments and their programs. Caution and continu­ous care on the part of extension service administrators are requiredto keep it so.

Experience and research have shown that the teaching approachesand procedures need continuous adaptation to the local culture. Thiscalls for a variety of extension teaching methods for practice,change, and adoption by individual farmers.

- _._--- - -------------------:

- 8 -

Continuous change in the approach to the people is necessary.T'nis is facilitated by what is sometimes called "grass roots" organi­zation of rural people. These are consciously planned, develGped andmaintained by organizational, leader, and member training. Coopera­tion with other agencies is important where research, academic educa­tion, and other public and private support services are separated.

Reference was made to the needs of rural families and theirinterests, and these are important in voluntary education. This,along with emphasis on learning by doing, is a motivating influence.The whole fa~ily involvement principle has a reinforcing effect onall aspects of social and agricultural development. Involving thewomen and girls is important because of their role in the laborforce as well as in family health and population levels.

The support of extension personnel who are specialists' in neededagriculture, home economics or social science areas is especiallyimportant for developing staff members who do not have strong basiceducation in these areas. These specialists'must be competent inextension methods too. The specialists, in turn, must be supportedby a knowledge source that is research based and focused on thepractical problems of agriculture and human development. z..t:>dernU. S. extension service sUbject-matter specialists generally are onpar in technical training with their research colleagues, a statusthat may be as important overseas as in our domestic programs.

Nr. Rice briefly mentions rural youth programs of the ExtensionService. Rural youth are a valuable resource of the future; parti­cularly in Latin American countries where such a large part of thepopulation is made up of persons under 25 years of age. Today privatesources may be doing more for rural youth than U. S. and foreigngovernments. The extension services have continued their 4-H-typeorganizations with significant results. During the 1960' s extensionservice youth programs in/Latin American and the Caribbean grew 250percent in membership. Enrollment in production projects by four outof every five members, along with group projects, has resulted inproduction valued at $5.5 million, and these are a major part of theyouth programs. With adequate U. S. and national government support,rural youth can be prepared for adulthood as efficient and productivefarm families.

In conclusion, we suggest that the foreign assistance goal todayshould be to fund and staff the U. S. assistance programs in waysthat will enable the existing weak and starved Latin American exten­sion services to reach maturity in the next decade or two. Con­currently, appropriate research sources and the needed auxiliaryinputs should be assured. Some of the strategies to accomplish thispurpose are suggested in Mr. Rice's study.

- 9 -

Footnotes

(1) Observations reported by J. Neil Raudabaugh made whileconducting a seminar for extension workers from certainBrazilian states.

(2) Nosher, A. T. "Creating A Progressive Rural structureto Serve A 140dern Agriculture," Agricultural Developmentcouncil, Inc., 630 Fifth Avenue, New York, New york 10020.(Pages 96-106; and 120-122.)

Remy T. FreireChief, Sectoral Studies Unit

Department of Economic AffairsOAS

The establishment of agricultural extension services in developingcountries seems to be representative of the United States bilateral foreignassistance program. Therefore, at this time of soul searching and reformu­lation of strategies among policy makers, Mr. Rice's study is timely. It isfortunate that he has carried out his assignement well. I am delighted tocomplement my detailed criticism of his draft with the following two generalcomments.

The first comment refers to the criteria for evaluating programs suchas agricultural extension. Mr. Rice adopted Mayer's scheme, which, from apurely technical point of view, seems to be quite satisfactory for thepurpose. However, if the "higher order" objectives of a technical assistanceprogram are what really matters, then the testing exercise has fallen short offulfilling its purpose. In fact the significance test, which is defined asasking "whether the project achieved the higher order objectives of thedevelopment programlt

, cannot be limited to the narrow aim of improving productionand productivity, no matter how desirable such improvements are.

Few people if any acquainted with both agriculture and development woulddeny the decisive importance of providing economic stimulus and opportunityfor farmers to produce more and better goods. Rice is quite correct inunderlining this fact. The parallel reasoning, that when such stimulus isnot present extension efforts are of little avail, could have been stressedfurther. This would help to explain why the correlation between such effortsand increases in output and productivity could hardly be significant in theabsence of the dominant economic determinants, or too small to be detectedwhen such determinants are present.

One could conclude, from the above, that the efforts to create viableextension services might well have been doomed from the beginning in allcases where the economic environment was not propitious. These were, by far,the majority. Yet the conclusion is valid only if we accept that the "highestorder't objective of the program was to increase production and productivity.It was not. Ultimately, the bilateral aid program must promote the interestsof the United States. In a great many external assistance projects, such asthe one under consideration, the United States interest is concretized into aneffort at image building, and the development of amicable political mutualrelations. Therefore, the valid test of significance should not have been thedegree of improvement in production and productivity but the kind of image andfeeling the United States effort left behind. In this respect it is notcertain that results could be measured accurately or that even if accurate,they would be necessarily valid in the longer run. Even Christ's mission onearth would have been rated insignificant immediately upon his death.

- 2 -

The second connnent concerns the so called "institutional building"effort. The supposition that the United States can go into other countriesand successfully inspire and/or build up institutions is a fallacy. Thetest of history has indicated time and again that institutions are its veryown for each country to build. To think otherwise has been one of the mostharmful misconceptions of the whole United States assistance effort abroad.At this time of uncertainty and concern, it would be wrong for a friend ofthe United States to belabor the point, but the sooner American power andingenuity are squarely placed at the service of projects in fields where theUnited States is truly outstanding -- and there are so very many, especiallyin the fields of technology -- and moves away from "institutional building",the better. In this sense, the fact that institutional building in agriculturalextension is being recognized as a failure, whether or not it is actuallytrue, is probably a blessing in disguise. Perhaps further institutionalbuilding efforts in other and much more delicate areas will not be hastilyundertaken.

~

,J

~t1;

f1'1j

iJ

~:?J.)

'?1

I,~

'~

i~

~

:~

J

~;~

-.4

:J

.f1\;;

•,'1!

,.

"

J

J,'11

~;j.

S'~

.•~

,~.~Jf'J,~1

Arthur T.' MOsherPresident

The Agricultural Development Council, Inc.

This study by Mr. Rice is obviously a very carefulattempt to evaluate the efforts of U.S. technical assistanceagencies to establish effective and viable agricultural ex­tension services in Latin America. The author concludes thatthese efforts were reasonably successful in establishing"viable" extension services (since most of them still exist)but not successful in establishing "effective" extensionservices, since in the areas he studied no positive correlationcould be found between intensity of extension activity andincreases in agricultural productivity •

He attributes the relative ineffectiveness of theseprograms to five factors. 'The two that he appears to considermost important are, first, a mistaken or inappropriate conceptof the role of extension and, second, inadequate concurrentattention to the essentials and the other accelerators of agri­cultural growthJl)I would agree with him with respect to bothof these factors. Throughout the period covered by Mr. Rice'sstudy, U.S. 90vernmental aid agencies tended to concentrate onwhat in the United States are govermnental functions -- researchand extension. Of these, extension received preponderant at­tention partly because of the mistaken assumption that the majorproblem was to get existing technologies disseminated andpartly because of the dependence of such agencies on annualappropriations. Simultaneously, the concept of the task ofextension as being almost exclusively educational was a reflec­tion of the role being played by the U.s. extension service inthe 1930's and 1940's, during which time those who went astechnicians to Latin America had started their professionalcareers. It was not until the late 50's and early 60's thatthe task of agricultural development began to be recognized asone requiring the simultaneous meeting of a number of require­ments including adequate market structures, distribution of farminputs and adequate economic incentives. Thus, during the periodwhen the programs studied by Mr. Rice were established there wasan exaggerated expectation of what extension alone could accomplish.

(1) A. T. Mosher, Getting Agriculture Moving, Frederick A.Praeger, Inc., New York, 1966

--------------_.....,_.. -- --

-2-

The author lists three secondary factors that alsolimited the effectiveness of the extension programs studied:inadequate training of extension workers, inadequate intensityof extension operations in particular areas, and inadequatebudgetary support. I would agree with the jUdgment cited bythe author that these limitations were so severe that no onecould have expected the extension efforts undertaken to besuccessful, no matter how well they might have been conceived.

Despite all of these shortcomings, the author con­cludes that there is an important role for lithe extensionprocess" to play and that the task now is to get it reorientedand to e,stablish effective linkages between it and the otheractivities that are essential to agricultural growth.

In approaching this problem, the author himself seemsto have a concept of extension and of its role that I considerinadequate. He seems to conceive the role of extension asbeing only to convey information to farmers. Instead, I wouldconceive the role of extension agents as being primarily tohelp farmers make the best use they can of the production op­portunities available to them, and secondarily to help farmerstake the initiative in getting the opportunities available tothem improved. (1)

I believe it is the author's limited concept of ex­tension that leads him to say "extension is a tool, and toolsought to be used by institutions but not institutionalized. II

If one adopts my broader definition of the role of extension,then it seems to me that it is important that it be undertakenby a ~eparate organization, even though I then agree with theauthor that the work of the extension organization should beclosely coordinated with that of other agencies operating inthe countryside to accelerate agricultural growth. I have out­lined a manner of accomplishing that coordination in a recentpUblication. (2) There is no point in establishing an extensionservice except in areas where there are one or more technologiesavailable that can sUbstantially increase production, adequate

(1) See the discussion of extension education on pages 532­535 of liThe World Food Problem, u.S. President's Science Ad­visory Committee, 11- Volume II.(2) A. T. Mosher, Creating a Progressive Rural Structure toServe a Modern Agriculture, Agricultural Development Council,1969.

•

-3-

facilities for distributing farm inputs, markets'for farmproducts, local verification trials to find the best packagesof agricultural practices under local conditions, productioncredit, and adequate rural transportation.

At several points the author comes out in favor ofimbedd.ing the extension process as he conceives it (takinginformation to farmers) within a fomento program as the solealternative to making extension the responsibility of aseparate organization. Where "fomento" is understood to meana comprehensive set of rural agri-support activities, I wouldagree. However, if by "fomento" is meant the promotion ofimproved technology with respect to a single crop, I thinkthere is a better alternative. The task of the individualfarmer is to make his own farm business as profitable as pos­sible and, in most cases, several different farm enterprisesare combined in the same farm business, or can be. Part ofthe task of helping each farmer make the best use of the pro­duction opportunities available to him (which to me is theprimary task of extension agents) is to help farmers decidebetween farm enterprises. Consequently, my own preference isfor having extension separately organized but grouped togetherwith the other agri-support activities with which it is sohighly complementary in a single "Division of Rural Agri­support Infrastructure" within the Ministry of Agr iculture,thus facilitating the coordination of all of these. Specialprograms to promote increased production of particular farmcommodities can then be formulated and coordinated by anOffice of Integrated Projects within the Ministry. (1)

At one point in his manuscript, Mr. Rice makes thestatement that the primary responsibility of an extension pro­gram is to work with those farmers who at one point he calls~'small farmers growing food crops" and at another point calls"subsistence farmers. II This also seems to me too narrow afocus. There is need for the extension process with all typesof farmers, even though the kinds of help that farmers mostneed vary with the size of farm businesses and with the natureof their production patterns.

(1) A.T. Mosher, Organizing and Planning to Create a ModernAgriculture, Lal Bahadur Shastri Memorial Lectures, IndianAgricultural Research Institute, March 1971

•

Delbert T. MYren.Chief

Planning Evaluation and Utilization DivisionOffice of A.I.D. Research and University Relations

Agency for International Development

Mr. Rice has asked that I focus my comments on the following majorconclusions of the study:

trA. The regular extension services' impact on farm productivity hasbeen very low.

trB. The reason for this is because extension worked independentlyof functionally related programs, trying to educate farmers toadopt new technology when there was no economic basis or supportfor adoption.

tre. This suggests that it was a mistake to try to institutionalizeextension as an independent authority: extension should be nomore than a tool in the hands of a higher authority.

r'D. The servicio environment was the wrong place to develop anextension service, because the inevitable problems associatedwith transfer only served to further isolate an already vulnerableoperation. tr

Let me say first of all that I am impressed by the scope of the author'sdetective work in reconstructing what happened. What bothers me is thepaucity of data. The period covered is not pre-history or even pre-Colombian,it is all within the past three decades. Yet the methods that the author hashad to use to uncover fragments of data remind one in many ways of thepainstaking work of the archeologist who reconstructs history from the bits ofpottery and organic waste left by early man. It is no criticism of this studybut certainly a relevant question here -- why do we not build into our actionprograms statements of goals and hypotheses and low-cost mechanisms forcollecting data to prove or disprove our hypotheses?/ Then each developmentexperience might serve as an experiment and yield much more information aboutsuccessful development methodology.

In respect to conclusion A, we can agree that the overall impact onproductivity has been low. In terms of returns per dollar of cost it's harderto judge the pay-off. Mr. Rice is aware of this and acknowledges that "thisis the central weakness of the study: its inability to quantify the amountof farm income, real and inputed, attributable to the efforts of the averageagent." At an average investment of $166,000 per country per year and anestimated cost of only $10,000 for operating a typical agency, the neededincreases in crop production to yield benefits above costs are very small. We

•... _- _. _._---- -----------------,

- 2 -

can argue this kind of detail but the author's conclusion is likely correct -­that what happened to farm productivity was not impressive.

Incidentally, Rice is saying nothing in these conclusions about secondgeneration problems nor that the wealthy gained more than the poor -- simplythat there was very little impact on productivity. Logically he does notdeal with distributive questions as these arise when there is an impact.

Accepting his conclusion as to low impact, the knowledge to be gainedfrom the study must be in terms of what can be learned about the reasons -­the subject matter of conclusions B, C, and D.

Let me approach this question by attempting to state a posteriori thehypotheses on which the action program apparently was based:

1. There exists a body of knowledge in the U.S., which if trans­ferred to farmers in Latin America could greatly increase productivity,

2. There exists an institution in the U.S. that if it were intro­duced to Latin America could bring about a transformation in agriculturalproduction and/or rural life,

3. That this knowledge could be transferred rapidly to LatinAmerican farmers by U.S. extension agents and agricultural informationspecialists,

4. That if these experts could be freed from budget and bureaucraticrestrictions, they could work more effectively,

5. Implied in all of the above, was a hypothesis that these wereeither sufficient conditions for bringing about the change or that all othernecessary conditions already existed.

This last point was of course the ultimate over-simplification. Creditwas not available, the inputs were not available locally, tenure conditionsworked against adoption, etc. Surprisingly, however, the other four points,all related specifically to knowledge transfer, also failed to be confirmedwhen tested in the field.

The know-how that the U.S. extension agents took to Latin America in the1950s is seen in a different sense in 1971. The new awareness has comeslowly and with some agony. Hardly any of the good answers worked. The soilswere different, the response to fertilizer was different, the insects weredifferent, the diseases were different, even the most desired plant characteristicswere different. Markets were unpredictable and the weather more so. In theend it was acknowledged that what the U.S. had to contribute was knowledge on

•

- 3 -

how to find answers rather than the answers themselves.

Hypothesis 2 on the positive impact of transplanting the U.S. extensionorganization was clearly rejected. U.S. extension 1950 style was a veryrefined creature with a body of doctrine to fit the state of development ofpost-war U.S. It was an organization that had gradually evolved to fitthose conditions. Its doctrine said that it should not discover newknowledge -- the task of research -- but only disseminate existing knowledge.Without adequate research networks, the model did not fit in Latin Americancountries.

Even the highly logical hypothesis 3 was found lacking. As mentioned,the agricultural problems were very different. Further the program wasstructured within the U.S. Government approach of short tours of duty. Thismeant that many of the experts never became proficient in Spanish and inthe majority of cases they moved before it became clear that the focus oftheir work would have to be changed if it were to be productive. The lackof Spanish and the roles of advisor and administrator (rather than fieldagent) both acted as barriers to a quick and accurate assessment of the realproblems.

As far as I can tell, hypothesis 4 gets neither support nor rejectionfrom the evidence at hand. If the content of the program had been adRquateI expect that this hypothesis might have been conf:irmed. If so, tho prublemsassociated with transfer into a government bureaucracy later would not havebeen sufficient reason for rejecting a highly successful approach. In factif the content of the program had been such that it was highly successful, Iquestion whether there would have been serious transfer problems.

One final comment on the question of training. It is possible that theearly training given in the U.S. in extension methodology may have contributedto the low impact. Even the later work in comparative extension methodsfocused rather narrowly on a defined methodology. Only in recent years hasthe awareness of these deficiencies led to efforts in various countries todevelop better locally based training. The most outstanding case is thegraduate program of the National School of Agriculture at Chapingo, Mexico.This two-year program which includes field work and research in the PueblaProject, was initiated in February 1969 with eight students from Mexico andfour other countries. Currently about 20 students are involved. The goal isto prepare a change agent with rural experience in a small farm setting inLatin America, who is able to conduct local field tests and who is trained notwith answers but with an understanding of how to confront a broad spectrum ofproblems related to increasing farm productivity and improving rural welfare.This kind of a program is the result of learning from the past experience ofothers in this field. And it offers a great deal of promise for more effective"extension" work in the future.

Let me handle the criticisms individually.

Remarks by the Author

The many points made by Matthews can be related to the four "conclusions!treported in the beginning of Myren's remarks. It is not clear whetherMatthews agrees with the first conclusion, that the impact on productivity hasbeen low. He shows that there has been no consensus on goals, and thattherefore productivity cannot be the sole test of significance. He pointsrightly to the poor data base available to the author. And he suggests thatthere could have been undetected increases in productivity which was con­sumed on the farm. This is doubtful, especially if purchased inputs wererequired, since subsistence farmers do not spend scarce funds to increase on­farm consumption except in emergencies. But he neither accepts nor deniesthe central conclusion about productivity.

We have no differences on the second conclusion, that serious problemshave arisen because of the failure of extension to link up with other ruralservices, and vice versa. There is a difference over the third conclusion,however, the finding that an extension service should be an instrument in thehands of a higher development authority. Matthews disputes this, saying thatextension has never been successfully planned at the national level, that theremust be grass-roots participation. But the author's point is that the extensionwork plan must be designed by persons not exclusively concerned with extension,not that it be designed at national headquarters rather than in the field. Theauthor doubts the validity of the grass-roots, program building, model too, butthat is another issue.

The fourth conclusion, regarding the difficulties inherent in the serV1ClOapproach to institution building, is not discussed in the otherwise exhaustiveMatthews statement.

Freire has his eye on issues that transcend the scope of the study, butgive cause to pause. It is not easy to accept either of his two major points(1) that U.S. interests may have been served whether or not viable institutionswere developed or improvements in productivity were achieved, and (2) thatinstitutional models developed in the U.S. cannot be deliberately recreatedwith U.S. help abroad, and therefore that AID shouldn't try. But the authorrecognizes that if they are true, the value of the study and the study con­clusions are less than is pretended.

The differences with Mosher are semantic. He uses the words "viable"and "effective" to mean something other than they mean in this study. Thelatter questions the viability of extension institutions, and thus gives theU.S. assistance program a low grade on effectiveness criteria. Mosher uses theword "extension" in a different way too. He refers to a type of extensionservice that he has been arguing for, and was perhaps applicable in an earlier

- 2 -

American period; I refer to the extension model that, unfortunately, tookshape in professional extension schools in the U.S. during the formativeperiod for extension in the study area and other parts of the developingworld. Finally, M::>sher questions the meaning of fomento, and sees the 1problem of fomento organizations narrowly conceived. The author failed inthe study to make clear that his fomento model includes any or all of the 1price, procurement and credit authorities. If M::>sher's proposed ministerial .structure includes these along with the other functions, and they are coordinated,then an assignment with him in such an institutional venture would be immenselysatisfying.

MYren says the same things the author tried to say in a better way. Hebrings up one of the central weaknesses of the study, which is restated belowalong with another weakness he and the other critics were too kind to expose.

(1) The efficiency analysis is inadequate. The author doesnot compute a benefit cost ratio for the average agency, yethe suggests that a credit service is more efficient than anextension service. Suppose a $10,000,000 AID loan for supervisedcredit were spent instead on financing a five year extensionbUdget. At $10,000 per agency (the author's figure), the loanwould finance an additional 200 agencies run in the conventionalway. If the costs per agency were doubled, so as to provide theagents with adequate training, transport, salaries and demonstrationalmaterial (no more than the incumbent directors argue they need),the loan would still allow the creation of 100 new agencies. Itwould triple the number of agencies in most countries. The authorhas not made the case that this would be a less efficient way toachieve rural progress than the supervised credit route. What hedoes show is that it would still be far less efficient than extensionprograms differently organized.

(2) The importance of markets is easily demonstrated. The creationof markets is not easily achieved, especially if these markets areto provide some "adequate" cash income for all farmers in an over­populated small holder economy; and if they are to involve basicfood crops. We may find that even a "well conceived lt

, extensionbased, rural development program may run out of markets. In thatcase the conventional extension services inadvertent "mistakelt ofturning rural youth to non-rural occupations may then appear to bewise.

In closing, it is necessary to point out that the critics were asked tocomment only on this condensed version of the general report. They did nothave access to the bulk of the productivity analysis. The author thinks thatsome of their doubts about the validity of the data base might be eased if theycould review that analysis. But he may be wrong.